EScholarship UC Item 97s2r3s1

Electronic Theses and Dissertations
UC San Diego
Peer Reviewed
Title:
Development of composite renewal systems for rapid rehabilitation and construction of bridge
decks
Author:
Pridmore, Anna Beth
Acceptance Date:
2009
Series:
UC San Diego Electronic Theses and Dissertations
Degree:
Ph. D., UC San Diego
Permalink:
https://escholarship.org/uc/item/97s2r3s1
Local Identifier:
b6288284
Abstract:
The deterioration of steel in aging reinforced concrete bridges is a continual problem which
could benefit from improved rehabilitation techniques that take advantage of enhanced and more
durable materials such as fiber reinforced polymer (FRP) composites. Appropriately designed
hybrid material systems benefit from the performance and durability advantages of FRP materials
yet remain more cost effective than comparable all-composite systems. Development of rapid
rehabilitation systems for the decks of concrete box girder bridges, which are increasingly
common throughout the United States, is presented. One goal of this research is to assess and
validate the use of FRP composite panels for use as both stay-in-place formwork and as the
bottom longitudinal and transverse reinforcement in the deck of concrete box girder bridges.
Performance assessments for full-scale two -cell box girder bridge specimens through monotonic
and extensive cyclic loading provided validation for the FRP panel system bridge deck as a viable
rehabilitation solution for box girder bridge decks. The FRP panel system performed comparably
to a conventionally reinforced concrete bridge deck in terms of serviceability, deflection profiles,
and system level structural interaction and performed superior to the RC bridge deck in terms of
residual deflections, and structural response under cyclic loading. Assessment of a damaged FRP
panel bridge deck system, which was repaired using a resin injection technique, showed superior
performance for the repaired system in terms of integrity of the FRP panel interface and cyclic
response. Rapid rehabilitation techniques for strengthening reinforced concrete box girder bridge
deck overhangs using near-surface-mounted (NSM) carbon fiber reinforced polymer (CFRP) were
also evaluated. Analytical predictions of load carrying capacity and deflections provided correlation
with experimental results, and the developed analysis methods provide an effective design tool for
future research. Results from the laboratory testing of a bridge deck overhang strengthened with
eScholarship provides open access, scholarly publishing

services to the University of California and delivers a dynamic
research platform to scholars worldwide.
FRP showed significant increases in load carrying capacity as well as deformation capacity as
compared to the as-built specimen without FRP. This research provides enhanced understanding
of hybrid structures and indicates significant potential for rehabilitation applications to concrete
box girder bridges.
Copyright Information:
All rights reserved unless otherwise indicated. Contact the author or original publisher for any
necessary permissions. eScholarship is not the copyright owner for deposited works. Learn more
at http://www.escholarship.org/help_copyright.html#reuse
eScholarship provides open access, scholarly publishing

services to the University of California and delivers a dynamic
research platform to scholars worldwide.
UNIVERSITY OF CALIFORNIA, SAN DIEGO
Development of Composite Renewal Systems for Rapid

Rehabilitation and Construction of Bridge Decks
A dissertation submitted in partial satisfaction of the
requirements for the degree Doctor of Philosophy
in
Structural Engineering
by
Anna Beth Pridmore
Committee in charge:
Professor Vistasp M. Karbhari, Chair
Professor Gilbert A. Hegemier, Co-Chair
Professor Prabhakar R. Bandaru
Professor Hyonny Kim
Professor Francesco Lanza di Scalea
Professor Vlado A. Lubarda
2009
Copyright
Anna Beth Pridmore, 2009
All rights reserved.
SIGNATURE PAGE
The dissertation of Anna Beth Pridmore is approved, and it is

acceptable in quality and form for publication on microfilm:
Co-Chair
Chair
University of California, San Diego

2009
iii
TABLE OF CONTENTS
Signature Page ...............................................................................................................iii
Table of Contents .......................................................................................................... iv
List of Abbreviations ...................................................................................................viii
List of Symbols............................................................................................................... x
List of Figures..............................................................................................................xiii
List of Tables .............................................................................................................. xxii
Acknowledgements ................................................................................................... xxiv
Vita ............................................................................................................................. xxv
Abstract...................................................................................................................... xxvi
INTRODUCTION .................................................................................................. 1
1.1 BACKGROUND AND MOTIVATION FOR RESEARCH ............................................ 1
1.2 FRP REHABILITATION OF CONCRETE BRIDGE DECKS....................................... 2
1.3 REHABILITATION NEEDS FOR REINFORCED CONCRETE BOX GIRDER BRIDGES 4
1.4 RESEARCH GOALS AND OBJECTIVES ................................................................. 6
1.5 SCOPE OF PROJECT............................................................................................ 7
LITERATURE REVIEW ....................................................................................... 8

2.1 OVERVIEW ........................................................................................................ 8
2.2 FORMWORK FOR CONSTRUCTION OF SLAB-ON-GIRDER BRIDGE DECKS ........... 9
2.2.1
Temporary Formwork versus Stay-in-Place Formwork ........................... 9
2.2.2
Stay-in-Place Formwork Made from Conventional Construction
Materials ............................................................................................................... 12
2.2.3
FRP Stay-in-Place Structural Formwork ................................................ 13
2.2.3.1
Composite Slabs Using FRP Panels with T-upstands ..................... 15
2.2.3.2
FRP Box Beams with Concrete in the Compression Zone ............. 20
2.2.3.3
FRP Panels with Rectangular Stiffeners ......................................... 25
2.3 TYPICAL REPAIR METHODS FOR REINFORCED CONCRETE BRIDGE DECKS ..... 32
2.4 NSM FRP SYSTEMS FOR FLEXURAL STRENGTHENING ................................... 34
2.4.1
NSM versus Externally Bonded FRP Reinforcement ............................ 34
2.4.2
Types of NSM Reinforcement................................................................ 36
2.4.3
NSM Failure Modes and Design Assumptions ...................................... 37
iv
2.4.3.1
Concrete crushing............................................................................ 37
2.4.3.2
FRP rupture ..................................................................................... 38
2.4.3.3
Bond failures ................................................................................... 38
2.4.3.4
Cover Splitting ................................................................................ 39
2.4.4
Prior Use ................................................................................................. 40
2.4.5
Available Specifications for Flexural Strengthening using NSM FRP .. 41
3
USE OF FRP PANELS AS THE STAY-IN-PLACE STRUCTURAL

FORMWORK SYSTEM FOR A BOX GIRDER BRIDGE DECK .................... 43
3.1 INTRODUCTION ............................................................................................... 43
3.2 OVERALL GEOMETRY AND CONSTRUCTION DETAILS ..................................... 43
3.3 FRP PANEL FEATURES .................................................................................... 54
3.4 PANEL PLACEMENT AND CONNECTION DETAILING ........................................ 58
3.5 SPECIMEN SETUP ............................................................................................ 62
3.6 INSTRUMENTATION ......................................................................................... 65
3.6.1
Linear Potentiometers and Inclinometers ............................................... 66
3.6.2
Strain Gages............................................................................................ 72
3.7 FORMWORK DEFLECTIONS DURING CASTING OF DECK CONCRETE................. 80
3.8 STATIC LOADING OF TEST SPECIMEN.............................................................. 82
3.8.1
Loading Protocol .................................................................................... 82
3.8.2
Overall Behavior .................................................................................... 84
3.8.3
Stiffness Change Throughout the Loading Range.................................. 87
3.8.4
Deflection Profiles .................................................................................. 95
3.8.5
Strain Development .............................................................................. 114
3.8.5.1
Comparison of Strains in FRP Panel Deck and RC Deck ............. 119
3.8.5.2
Strain Profiles along the Longitudinal Direction of the Deck ....... 127
3.8.6
Strain Progression and Maximum Strain Levels .................................. 131
3.8.7
Unloading Response ............................................................................. 133
3.8.7.1
Loading versus unloading stiffness comparisons .......................... 133
3.8.7.2
Residual deflections and strains .................................................... 136
3.8.8
Separation between FRP Panels and Concrete ..................................... 138
3.9 LIMIT STATES ............................................................................................... 143
3.10 DESIGN THRESHOLDS ................................................................................... 146
3.11 CYCLIC LOADING OF TEST SPECIMEN ........................................................... 147
3.11.1 Loading Protocol .................................................................................. 147
3.11.2 Changes in Center Deflections during Cyclic Loading ........................ 149
3.11.3 Changes in Strains during Cyclic Loading ........................................... 153
3.11.4 Progression of Separation between FRP Panel and Concrete .............. 157
3.12 SUMMARY..................................................................................................... 160
REPAIR OF THE FRP PANEL STRUCTURAL FORMWORK SYSTEM ..... 162

4.1 INTRODUCTION ............................................................................................. 162
4.2 DESIGN AND IMPLEMENTATION OF REPAIR ................................................... 163
INSTRUMENTATION.................................................................................................. 167
4.3 RESULTS FOR STATIC LOADING OF THE REPAIRED SPECIMEN ....................... 168
v
4.3.1
Loading Protocol .................................................................................. 168
4.3.2
Overall Behavior .................................................................................. 169
4.3.3
Stiffness Change for Repaired FRP Panel Side .................................... 173
4.3.4
Deflection Profiles ................................................................................ 177
4.3.5
4.3.5.1
Strain Progression for Gages at Center of Composite Side Deck . 189
4.3.5.2
Strain Profiles along the Longitudinal Direction of the Deck ....... 197
4.3.6
4.4 RESULTS FOR CYCLIC LOADING OF THE REPAIRED SPECIMEN ...................... 204
4.4.1
4.4.2
Changes in Center Deflections during Cyclic Loading ........................ 205
4.4.3
Changes in Strains during Cyclic Loading ........................................... 208
4.4.4
Changes in Separation between the FRP Panel and the Concrete ........ 210
4.4.5
Static Loading Following the Repaired Specimen Cyclic Phase ......... 212
4.5 SUMMARY..................................................................................................... 212
5
CYCLIC RESPONSE OF THE FRP PANEL STAY-IN-PLACE STRUCTURAL

FORMWORK SYSTEM .................................................................................... 215
5.1 INTRODUCTION ............................................................................................. 215
5.2 OVERALL GEOMETRY AND CONSTRUCTION DETAILS ................................... 215
5.3 PANEL PLACEMENT AND CONNECTION DETAILING ...................................... 218
5.4 SPECIMEN SETUP AND INSTRUMENTATION ................................................... 226
5.4.1
Linear Potentiometers and Inclinometers ............................................. 227
5.4.2
Strain Gages.......................................................................................... 230
5.5 DATA FROM CASTING OF DECK CONCRETE .................................................. 238
5.6 STATIC LOADING OF TEST SPECIMEN............................................................ 241
5.6.1
5.6.2
Overall Behavior .................................................................................. 242
5.6.3
Stiffness change per increment ............................................................. 245
5.6.4
5.6.5
5.6.5.1
Comparison of Strains in FRP Panel Deck and RC Deck ............. 255
5.6.5.2
Strain Profiles along the Longitudinal Direction of the Decks ..... 261
5.6.6
Strain Progression and Maximum Strain Levels .................................. 263
5.6.7
Residual Deflections and Strains .......................................................... 265
5.6.8
5.7 CYCLIC LOADING OF TEST SPECIMEN ........................................................... 269
5.7.1
5.7.2
Crack Patterns ....................................................................................... 271
5.7.3
Progression of Center Deflections ........................................................ 271
5.7.4
5.7.5
Summary............................................................................................... 284
USE OF NEAR SURFACE MOUNTED FRP REINFORCEMENT AS A

MEANS OF RAPID BRIDGE DECK REHABILITATION ............................. 286
vi
6.1 PROJECT SPECIFIC NEED FOR FRP REHABILITATION .................................... 286

6.2 SPECIMEN GEOMETRY AND CONSTRUCTION ................................................. 288
6.3 LOADING SETUP............................................................................................ 289
6.4 AS-BUILT TEST ............................................................................................. 290
6.4.1
Demand Calculations............................................................................ 291
6.4.2
Capacity Calculations ........................................................................... 293
6.4.3
Analytical Predictions .......................................................................... 296
6.4.4
Instrumentation ..................................................................................... 298
6.4.5
6.4.6
Experimental Results ............................................................................ 302
6.4.7
Comparison with Theory ...................................................................... 311
6.5 REHABILITATED TEST ................................................................................... 312
6.5.1
Calculations for Potential CFRP NSM Strengthening Schemes .......... 313
6.5.2
Analytical Strength ............................................................................... 314
6.5.3
Options for Rehabilitation .................................................................... 315
6.5.4
Rehabilitation Construction .................................................................. 322
6.5.5
Capacity Calculations ........................................................................... 327
6.5.6
Instrumentation ..................................................................................... 327
6.5.7
6.5.8
Experimental Results ............................................................................ 334
6.5.9
Comparison with Theory ...................................................................... 346
6.6 COMPARISON OF FRP STRENGTHENED AND AS-BUILT OVERHANGS ............ 348
6.7 SUMMARY OF RESULTS ................................................................................. 355
7
CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK ....... 356

7.1 SUMMARY AND CONCLUSIONS ..................................................................... 356
7.1.1
FRP Panel SIP Structural Formwork System for Box Girder Bridges . 356
7.1.2
Rehabilitation of Bridge Decks Using Near Surface Mounted FRP .... 359
7.2 CONCLUSIONS ............................................................................................... 360
7.2.1
7.2.2
7.3 RECOMMENDATIONS FOR FUTURE RESEARCH .............................................. 362
7.3.1
7.3.2
REFERENCES ........................................................................................................... 365
vii
LIST OF ABBREVIATIONS
AASHTO
American Association of State Highway and Transportation Officials,
ACI
American Concrete Institute,
AFRP
Aramid Fiber Reinforced Polymer,
ASTM
American Society for Testing and Materials,
CFRP
Carbon Fiber Reinforced Polymer,
CHBDC
Canadian Highway Bridge Design Code,
DSC
Differential Scanning Calorimetry,
DMTA
Dynamic Mechanical Thermal Analysis,
Degrees Celsius,
Degrees Fahrenheit,
ft
Feet,
FRP
Fiber Reinforced Polymer,
GFRP
Glass Fiber Reinforced Polymer,
GPa
Giga-Pascals (109 Pascals),
hwy
Highway,
in
Inches,
hwy
Highway,
One thousand,
kip
Kilo-pound (103 pounds)
ksi
Kilo-pounds per square inch (103 pounds per square inch),
kN
Kilo-Newtons (103 Newtons),
LC
Lightweight Concrete,
LRFD
Load and Resistance Factor Design,
mil
Million,
mm
Millimeters,
Meters,
Microstrains,
MPa
Mega-Pascals (106 Pascals),
Msi
Mega-pounds per square inch (103 pounds per square inch),
viii
Newtons,
NC
Normal weight concrete
NSM
Near Surface Mounted,
NSM
Near Surface Mounted Reinforcement,
o.c.
On Center,
pot
Linear Potentiometer
psf
Pounds per Square Foot,
rad.
Radians,
RC
Reinforced Concrete,
SIP
Stay-in-place,
TYP
Typical,
UCSD
University of California, San Diego,
US
United States,
ix
LIST OF SYMBOLS
a
Depth of concrete compression block, assuming a rectangular stress

distribution,
Ac
Area of net concrete section,
As
Area of tensile steel reinforcement,
Ag
Reduction coefficient that accounts for level of aggregate coverage of

the top surface of FRP deck form, recommended by Dieter as addition
to ACI shear capacity equation,
Section width,
bw
Width of the concrete slab,
CE
Environmental reduction factor defined in ACI440.2R Table 8.1,
Distance from the extreme compression fiber to tensile reinforcement,
df
Distance from the compression fiber to the centroid of the FRP
d FRP
Distance from extreme compression fiber to bottom of FRP panel
Deflection at target actuator load, Pf , at the location under

consideration,
Deflection at minimum threshold actuator load, Pi , at the location

under consideration,
Deflection at nominal zero actuator load, Po , at the location under

consideration,
max_
form
Maximum allowable deflection of stay-in-place formwork,
Ec
Youngs modulus of concrete,
Ef
Experimentally determined modulus of elasticity of FRP
EL
Longitudinal modulus of FRP composites,
E LF
Experimental flexural modulus of FRP composites in longitudinal

direction,
Es
Youngs modulus of steel,
ET
Transverse modulus of FRP composites,
cu
Maximum usable strain of concrete,
Strain level in FRP reinforcement,
fe
Effective strain in FRP reinforcement,
fu
Design rupture strain in FRP reinforcement,
Strain in steel,
s*
Adjusted steel strains,
f c'
Specified compressive strength of concrete at 28 days,
ffe
Effective stress in the FRP assuming elastic behavior,
f fu
Design ultimate tensile strength of the FRP reinforcement,
f fu*
Ultimate tensile strength of the FRP as reported by the manufacturer,
fr
Modulus of rupture of concrete,
fy
Yield stress of steel,
G LH
In-plane shear modulus of FRP composites,
Density,
Additional reduction factor from ACI 440.06R (Section 9.6.1),
Ic
Second moment of inertia of net concrete section about its centroid,
I cr
Moment of inertia of cracked section transformed to concrete,
Ie
Effective moment of inertia,
Ig
Gross moment of inertia,
Effective length factor for compression members (Equation 3.6),
Span length of the deck measured from center-to-center of supports,
Lspan
Span length of the stay-in-place formwork,
M cr
Cracking moment,
Mn
Moment capacity of the slab,
Mu
Factored moment in the slab at the edge of the outer vertical stem
adjacent to the overhang,
xi
Modular ratio of steel to concrete,
Pf
Target actuator load of a load cycle based on loading protocol,
Pi
Minimum threshold actuator load of a load cycle,
Po
Nominal zero actuator load at the start of experimental testing,
Curvature in deck slab (Equation 3.7),
LF
Longitudinal flexural strength of FRP composites,
TF
Transverse flexural strength of FRP composites,
LT
Longitudinal tensile strength of FRP composites,
TT
Transverse tensile strength of FRP composites,
Resistance factor,
Reinforcement ratio,
slim it
Maximum spacing limit for FRP,
Thickness of deck slab,
tL
Thickness of the near surface mounted CFRP strips,
Interlaminar strength of FRP composites,
Vc
Shear capacity of the slab,
Vu
Factored shear in the slab at the edge of the outer vertical stem adjacent
to the overhang,
weight wall
The total load applied to the overhang due to the combined weight of
the sound wall and the traffic barrier,
yt
Distance from the neutral axis to the extreme tensile fiber,
xii
LIST OF FIGURES
Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
Conventional wooden shored formwork layout for deck (after [76]) ........ 10
Concrete filled membrane deck system (after [95])................................... 14
Composite slab with structural formwork (after [96]) ............................... 15
Beam specimens with FRP SIP forms (after [73]) ..................................... 16
Sandwich panel with a lightweight concrete core, concrete in
compression zone, and FRP panel with T-upstands in tension zone
[97, 98] .................................................................................................... 17
Figure 2.6 Use of commercially available FRP plank as SIP forwork and
secondary reinforcement in reinforcement-free deck [66] ...................... 19
Figure 2.7 FRP box beam with concrete in compression zone (after [100]) ............... 21
Figure 2.8 FRP box beam with concrete compression flange (after [106]) ................ 22
Figure 2.9 Hollow trapezoidal tubes partially filled with concrete (after [108]) ........ 23
Figure 2.10 Corrugated GFRP SIP formwork acting as reinforcement for a
concrete beam (after [79]) ....................................................................... 24
Figure 2.11 Cross section of hybrid FRP panel-concrete bridge deck system
with SIP formwork and tubular stiffeners used in the Salem
Avenue Bridge [113] ............................................................................... 26
Figure 2.12 Hybrid system used in US-151/ Hwy 26 Bridge consisting of
pultruded FRP structural form with aggregate bonded surface and
tubular stiffners as well as FRP grid for top layer of reinforcement
(after [67]) ............................................................................................... 27
Figure 2.13 Steel-free hybrid FRP-concrete bridge with FRP tubular girders
[119] ........................................................................................................ 29
Figure 2.14 Steel-free hybrid FRP-concrete bridge with FRP tubular girders
[119] ........................................................................................................ 30
Figure 2.15 Rehabilitation of bridge deck slabs using externally bonded FRP
reinforcement [128] ................................................................................. 35
Figure 2.16 Bond failure modes for FRP strips (after [134]) ...................................... 39
Figure 3.1 Overall representation of test specimen ..................................................... 44
Figure 3.2 Overall test specimen dimensions- front elevation view ........................... 46
Figure 3.3 Top layer of steel reinforcement for deck slab- plan view......................... 47
Figure 3.4 Steel reinforcement detailing in lower portion of box girder - front
elevation view .......................................................................................... 48
Figure 3.5 Steel reinforcement detailing in lower portion of box girderside
elevation view .......................................................................................... 49
Figure 3.6 Lower portion of the box girder bridge before and after casting of
concrete.................................................................................................... 51
Figure 3.7 Strength development for concrete from 1st pour ..................................... 52
Figure 3.8 Strength development for concrete from 2nd pour (deck) ......................... 52
Figure 3.9 Construction of test specimen formwork for 2nd pour .............................. 53
Figure 3.10 Pouring of deck concrete and finishing of deck surface .......................... 53
Figure 3.11 FRP deck panel system [119]................................................................... 54
xiii
Figure 3.12
Figure 3.13
Figure 3.14
Figure 3.15
Figure 3.16
Figure 3.17
Figure 3.18
Figure 3.19
Figure 3.20
Figure 3.21
Figure 3.22
Figure 3.23
Figure 3.24
Figure 3.25
Figure 3.26
Figure 3.27
Figure 3.28
Figure 3.29
Figure 3.30
Figure 3.31
Figure 3.32
Figure 3.33
Figure 3.34
Figure 3.35
Figure 3.36
Figure 3.37
Figure 3.38
Figure 3.39
Figure 3.40
Figure 3.41
Figure 3.42
Figure 3.43
Figure 3.44
Cross-section details of the stiffeners on the deck panel (after [120])..... 55

Schematic of FRP panel locations ........................................................... 59
FRP panel installation details ................................................................... 60
Plane view looking down on reinforcement system in deck, as built ...... 61
Load cell placement below test specimen ................................................ 62
Detail of load cell used to support test specimen ..................................... 63
Overall specimen setup ............................................................................ 64
Layout of linear potentiometers located on the underside of the
deck.......................................................................................................... 68
Layout of linear potentiometers located on the top of the deck ............... 69
Detail of horizontally oriented linear potentiometer and
inclinometer ............................................................................................. 70
Layout of inclinometers and horizontally mounted linear
potentiometers ......................................................................................... 71
Strain gage pattern locations for gages attached to vertical stem
rebar ......................................................................................................... 72
Strain gage pattern A ............................................................................ 73
Strain gage pattern B ............................................................................ 73
Layout of longitudinal strain gages attached to bottom of FRP side
deck.......................................................................................................... 74
Layout of transverse strain gages attached to bottom of FRP side
deck.......................................................................................................... 75
Layout of strain gages attached to top longitudinal steel
reinforcement ........................................................................................... 76
Layout of strain gages attached to bottom longitudinal steel
reinforcement ........................................................................................... 77
Layout of strain gages attached to top transverse steel
reinforcement ........................................................................................... 78
Layout of strain gages attached to bottom transverse steel
reinforcement ........................................................................................... 79
Comparison of formwork center deflections under construction
loading ..................................................................................................... 80
Envelope plot of load versus center deflections....................................... 85
Load versus center deflections ................................................................. 86
Effective stiffness ratios of the two slabs determined from linear
potentiometer data ................................................................................... 89
Illustration showing difference in stiffness calculation methods ............. 93
Effective stiffness ratios for two methods of calculating stiffness ......... 94
Deflection profile along length of deck above central stem .................... 96
Deflection profile along length of deck above composite side stem ....... 97
Deflection profile along length of deck above RC side stem .................. 97
Deflection profile along centerline of composite side deck ..................... 99
Deflection profile along centerline of RC side deck ................................ 99
Comparison of deflection profiles at a load level of 756 kN ................. 100
Comparison of deflection profiles at a load level of 845 kN ................. 101
xiv
Figure 3.45 Comparison of deflection profiles at ultimate capacity (890 kN) .......... 101
Figure 3.46 Percent change between composite side and RC side deflection
values at different load levels- linear potentiometers located at 2/3
span along the centerline of each deck ................................................ 103
values at different load levels- linear potentiometers located at 1/3
span along the centerline of each deck ................................................. 104
Figure 3.48 Percent change between composite side and reinforced concrete
side deflection values at different load levels- linear potentiometers
located at 1/3 span along the midspan of each deck ............................ 104
values at different load levels- linear potentiometers located at the
center of each deck ................................................................................ 105
Figure 3.50 Deflection profile along row 2 of linear potentiometers ........................ 106
Figure 3.54 Deflection profile along centerline of specimen .................................... 109
Figure 3.55 Deflection profile along row A of linear potentiometers ....................... 110
Figure 3.56 Deflection profile along row B of linear potentiometers ....................... 110
Figure 3.57 Deflection profile along row C of linear potentiometers ....................... 111
Figure 3.58 Envelope plot of load versus deflections for linear potentiometers
along the centerlines of the two slabs .................................................... 112
Figure 3.59 Envelope plot of load versus deflections for linear potentiometers
along the centerline of the test specimen ............................................... 113
Figure 3.60 Deck slab cross-section illustrating location of strain gages within
depth ...................................................................................................... 115
Figure 3.61 Illustration of strain distribution and adjustment of RC side strains...... 116
Figure 3.62 Assumed linear stress distribution in straight line theory ...................... 118
Figure 3.63 Illustration of strain distribution and adjustment of RC side strains...... 119
Figure 3.64 Load versus strain for bottom longitudinal strain gages ........................ 120
Figure 3.65 Load versus strain for bottom transverse strain gages ........................... 121
Figure 3.66 Strain profile for top transverse rebar along centerline.......................... 123
Figure 3.67 Load versus strain for top transverse strain gages ................................. 124
Figure 3.68 Load versus strain for top transverse strain gages near central stem ..... 125
Figure 3.69 Load versus strain for top transverse strain gages near outer stems ...... 126
Figure 3.70 Strain profile for gages attached to bottom of FRP panel along line
3 (Line of gages is located directly along the composite side
centerline) .............................................................................................. 129
Figure 3.71 Strain profile for bottom longitudinal rebar along line 17 ..................... 129
Figure 3.72 Strain profile for top longitudinal rebar along line 8 ............................. 130
Figure 3.73 Stiffness ratios of the center of the two slabs found from deflection
data (loading portion data was recalculated with 222kN instead of
98 kN as the baseline)............................................................................ 136
xv
Figure 3.74 Representative load versus strain plot to illustrate difference in

residual strains between composite side and RC side strain gages. ...... 138
Figure 3.75 Load versus deflection plots for detection of FRP panel separation...... 140
Figure 3.76 Load versus deflection plots for detection of FRP panel separation...... 140
Figure 3.77 Load versus strain comparison of three transverse gages along test
specimen centerline at the same location and at different depths
within the deck ...................................................................................... 142
within the deck ...................................................................................... 142
Figure 3.79 Effective stiffness versus cycle number for phase II of cyclic
loading ................................................................................................... 152
Figure 4.1 Region in which the FRP panels had separated from the concrete .......... 163
Figure 4.2 Detail of hole pattern arrangement for repair of FRP side deck .............. 165
Figure 4.3 Injection port in top of deck and setup used to inject resin into deck ...... 166
Figure 4.4 Load versus deflection envelope plots of the original and repaired
test specimen for both the FRP and RC slabs........................................ 170
Figure 4.5 Load versus deflection envelope plots of the original and repaired
test specimen for both the FRP and RC slabs........................................ 171
Figure 4.6 Stiffness ratios of the FRP side slab at different loading phases ............. 175
Figure 4.7 Deflection profile of repaired specimen on the deck above the central
stem........................................................................................................ 178
Figure 4.8 Deflection profile along centerline of original FRP side deck ................. 179
Figure 4.9 Deflection profile along centerline of repaired FRP side deck ................ 179
Figure 4.10 Original and repaired FRP side deflection profiles at 400 kN load
level ....................................................................................................... 180
level ....................................................................................................... 181
level ....................................................................................................... 181
Figure 4.13 Transverse deflection profile for original composite side slab .............. 183
Figure 4.14 Transverse deflection profile for repaired composite side slab ............. 183
level ....................................................................................................... 184
level ....................................................................................................... 185
Figure 4.17 Envelope plot for linear potentiometers along the centerlines of the
original and repaired FRP slab .............................................................. 186
Figure 4.18 Envelope plot for linear potentiometers on original and repaired
FRP side along the centerline of the test specimen ............................... 187
Figure 4.19 Original and repaired test specimen load versus strain plots for the
bottom longitudinal FRP side strain gage, CM3L ................................. 191
bottom transverse FRP side strain gage, CM3T .................................... 193
xvi
top longitudinal FRP side strain gage, DTL8-3..................................... 195
top transverse FRP side strain gage, DTT10-3 ...................................... 196
Figure 4.23 Original specimen strain profile for gages attached to bottom of
FRP panel along line 3 .......................................................................... 198
Figure 4.24 Repaired specimen strain profile for gages attached to bottom of
FRP panel along line 3 .......................................................................... 198
Figure 4.25 Strain profile for original specimens top longitudinal rebar along
line 8 ...................................................................................................... 200
Figure 4.26 Strain profile for repaired specimens top longitudinal rebar along
line 8 ...................................................................................................... 200
Figure 4.27 Load versus deflection plots for linear potentiometers on top and
bottom of repaired FRP side deck ......................................................... 202
bottom of repaired FRP side deck ......................................................... 203
Figure 4.29 Effective stiffness versus cycle # for original and repaired FRP side
deck........................................................................................................ 206
Figure 4.30 Envelope plot of load versus deflections of repaired FRP side
before and after cyclic loading phase .................................................... 212
Figure 5.1 Strength development for deck slab concrete in test specimen #2 .......... 217
Figure 5.2 Construction of test specimen formwork for 2nd pour ............................ 217
Figure 5.3 Schematic of FRP panel locations (same overall locations for both 1st
and 2nd test specimens) .......................................................................... 219
Figure 5.4 End connection detailing of FRP panels .................................................. 220
Figure 5.5 Drilling of stem concrete using FRP panels for alignment of holes ........ 221
Figure 5.6 Detail of anchor position with respect to concrete stem and FRP
panel ...................................................................................................... 223
Figure 5.7 Composite side deck after FRP panel installation ................................... 225
Figure 5.8 Reinforcement system in deck, as built ................................................... 225
Figure 5.9 Test specimen with instrumentation, prior to installation of actuators .... 226
Figure 5.10 Layout of linear potentiometers located on the underside of the
composite deck ...................................................................................... 228
Figure 5.11 Layout of linear potentiometers located on the top of the composite
deck........................................................................................................ 229
Figure 5.12 Layout of longitudinal strain gages attached to bottom of FRP side
deck........................................................................................................ 231
Figure 5.13 Layout of transverse strain gages attached to bottom of FRP side
deck........................................................................................................ 232
Figure 5.14 Layout of longitudinal strain gages attached to top of central FRP
panel ...................................................................................................... 233
Figure 5.15 Layout of transverse strain gages attached to top of central FRP
panel ...................................................................................................... 233
Figure 5.16 Layout of strain gages attached to top longitudinal steel
reinforcement ......................................................................................... 234
xvii
Figure 5.17 Layout of strain gages attached to top transverse steel

reinforcement ......................................................................................... 235
Figure 5.18 Layout of strain gages attached to bottom transverse steel
reinforcement ......................................................................................... 236
Figure 5.19 Strain gage pattern locations for gages attached to vertical stem
rebar ....................................................................................................... 237
Figure 5.20 Strain gage pattern A .......................................................................... 238
Figure 5.21 Strain gage pattern B .......................................................................... 238
Figure 5.22 Comparison of formwork center deflections under construction
loading ................................................................................................... 239
Figure 5.23 Temperature progression over three days for freshly poured deck
concrete.................................................................................................. 240
Figure 5.24 Load versus center deflections ............................................................... 244
Figure 5.25. Envelope plot of load versus center deflections.................................... 245
Figure 5.26 Effective stiffness ratios of the two slabs............................................... 247
Figure 5.27 Deflection profile along length of deck above central stem .................. 249
Figure 5.28 Deflection profile along centerline of the composite side deck ............. 250
Figure 5.29 Deflection profile along centerline of the RC side deck ........................ 250
Figure 5.30 Comparison of deflection profiles at a load level of 578 kN ................. 251
Figure 5.33 Deflection profile along centerline of specimen .................................... 254
Figure 5.34 Load versus strain for bottom longitudinal strain gages ........................ 256
Figure 5.35 Load versus strain for bottom transverse strain gages ........................... 257
Figure 5.36 Strain profile for top transverse rebar along centerline.......................... 258
Figure 5.37 Load versus strain for top transverse strain gages ................................. 259
Figure 5.38 Load versus strain for top transverse strain gages near central stem ..... 260
Figure 5.39 Load versus strain for top transverse strain gages above outer stems ... 260
Figure 5.40 Strain profile for gages attached to bottom of FRP panel along line
3 (Line of gages is located directly along the composite side
centerline) .............................................................................................. 262
Figure 5.41 Strain profile for bottom longitudinal rebar along line 17 ..................... 262
bottom of deck- located along test specimen centerline and 610 mm
away from the composite side centerline, 178 mm from the
composite side edge of the center stem ................................................. 267
within the deck ...................................................................................... 268
within the deck ...................................................................................... 268
Figure 5.45 Loading phases for test specimen #2 ..................................................... 270
Figure 5.46 FRP side center deflections throughout cyclic loading phases .............. 272
Figure 5.47 FRP panel side center deflections throughout cyclic loading phases .... 273
xviii
Figure 5.48
Figure 5.49
Figure 5.50
Figure 5.51
Effective stiffness ratios of the two slabs throughout cyclic loading .... 275
Center deflections at 1x service load throughout cyclic loading ........... 276
Residual deflection comparisons throughout cyclic loading ................. 277
Deflection response at 2x service load along centerline during
cyclic loading......................................................................................... 279
Figure 5.52 Deflection response of FRP side at 2x service load during cyclic
loading ................................................................................................... 281
Figure 5.53 Deflection response of RC side at 2x service load during cyclic
loading ................................................................................................... 281
Figure 5.54 Positioning of linear potentiometers used for deflection
comparisons ........................................................................................... 282
Figure 5.55 Composite side and RC side deflections at different cyclic loading
stages ..................................................................................................... 282
stages ..................................................................................................... 283
stages ..................................................................................................... 283
stages ..................................................................................................... 284
Figure 6.1 Sound wall placed onto the overhang of a box girder bridge deck .......... 286
Figure 6.2 Overall dimensions of specimen .............................................................. 289
Figure 6.3 Plan view of test fixture used to load overhang ....................................... 290
Figure 6.4 Overall test setup ...................................................................................... 290
Figure 6.5 Standard CALTRANS masonry sound wall design [159] ....................... 292
Figure 6.6 Moment-curvature response for as-built specimen .................................. 295
Figure 6.7 Adjustment of moment of inertia along moment curvature profile ........ 297
Figure 6.8 Moment profile on deck due to a point load at the edge of the
overhang ................................................................................................ 298
Figure 6.9 Instrumentation locations used for measuring compression of bearing
pad ......................................................................................................... 299
Figure 6.10 Position of linear potentiometers for measuring deflections of deck
slab (Note: not to scale) ......................................................................... 300
Figure 6.11 Specific test setup schematic for as-built specimen ............................... 301
Figure 6.12 Deck slab prior to experimental testing ................................................. 303
Figure 6.13 Initial cracking of deck slab at 84kN (19 kip) per jack- top view of
deck........................................................................................................ 304
Figure 6.14 Initial cracking of deck slab at 84kN (19 kip) per jack - side view of
deck........................................................................................................ 304
Figure 6.15 Crack marking of deck slab at 102 kN (23 kip) per jack- top view of
deck........................................................................................................ 305
Figure 6.16 Cracking observed at ultimate capacity- side view of deck ................... 305
Figure 6.17 Cracking observed at ultimate capacity- top view of deck .................... 306
Figure 6.18 Detail of cracking at ultimate capacity in central section of deck
near loading beam.................................................................................. 306
Figure 6.19 Cracks observed on top of slab tested to ultimate capacity ................... 306
xix
Figure 6.20 Deck slab tested to ultimate capacity after removal of loose
concrete.................................................................................................. 307
Figure 6.21 Detail of most severely damaged section ............................................... 307
Figure 6.22 Comparisons of deflections at the edge of the deck slab overhang ....... 308
Figure 6.23 Deflection profile along the outer edge of specimen (Line A) .............. 309
Figure 6.24 Deflection profile along center of specimen (Line M) .......................... 309
Figure 6.25 Deflection profile along the central edge of specimen (Line B) ............ 310
Figure 6.26 Deflections midway along overhang (Line 3)........................................ 310
Figure 6.27 Deflections directly below loading beam (Line 4)................................. 311
Figure 6.28 Deflection values from experimental results with analytical
predictions ............................................................................................. 312
Figure 6.29 Plan view of deck illustrating chosen CFRP strengthening scheme ...... 320
Figure 6.30 Grooves cut in deck for NSM strengthening ......................................... 322
Figure 6.31 Application of resin system used in grooves to bond CFRP strips to
concrete.................................................................................................. 324
Figure 6.32 CFRP strips installed .............................................................................. 325
Figure 6.33 Progression of Tg determined from DMTA tests ................................... 326
Figure 6.34 Progression of Tg determined from DSC tests ....................................... 326
Figure 6.35 Position of linear potentiometers for measuring deflections of deck
slab ......................................................................................................... 329
Figure 6.36 Strain gage patterns and designations .................................................... 330
Figure 6.37 Position of strain gages attached to CFRP strips ................................... 330
Figure 6.38 Completed installation of CFRP strips with full instrumentation
setup ....................................................................................................... 331
Figure 6.39 Side view of completed installation of CFRP strips with full
instrumentation setup............................................................................. 331
Figure 6.40 Cracking of deck slab at 184 kN (41 kip) per jack- top view of deck ... 335
Figure 6.41 Cracking of deck slab at 184 kN (41kip) per jack - side view of
deck........................................................................................................ 335
Figure 6.42 Cracking observed at ultimate capacity- side view of deck ................... 336
Figure 6.43 Cracking of deck slab at ultimate capacity- top view of deck ............... 336
Figure 6.44 Side view of tested FRP rehabilitated specimen after removal of
loose concrete ........................................................................................ 337
Figure 6.45 Comparisons of deflections at the edge of the deck slab overhang ....... 338
Figure 6.46 Deflection profile along the central edge of specimen (Line B) ............ 339
Figure 6.47 Deflection profile along center of specimen (Line M) .......................... 339
Figure 6.48 Deflection profile along the outer edge of specimen (Line A) .............. 339
Figure 6.49 Deflections midway along overhang (Line 3)........................................ 340
Figure 6.50 Deflections directly below loading beam (Line 4)................................. 340
Figure 6.51 Strain profile along the central edge of specimen (Line 9) .................... 341
Figure 6.52 Strain profile along the middle of specimen (Line 5) ............................ 342
Figure 6.53 Strain profile along the outer edge of specimen (Line 1) ...................... 342
Figure 6.54 Strains along the edge of the stem adjacent to the deck overhang
(Line B) ................................................................................................. 343
xx
Figure 6.55 Shear stress distribution within adhesive along the middle of
specimen (Line 5) .................................................................................. 344
Figure 6.56 Location of cut for removal of FRP strengthened overhang ................. 345
Figure 6.57 Detail showing actual location of reinforcement ................................... 345
Figure 6.58 Detail of failure surface of FRP strip ..................................................... 346
Figure 6.59 Experimental and analytical load versus deflection plots for FRP
strengthened deck overhang .................................................................. 348
Figure 6.60 Comparison of load versus deflection profiles for deck overhang ........ 349
Figure 6.61 Comparisons of deflections of the deck overhang at different load
levels ...................................................................................................... 350
Figure 6.62 Comparison of center deflection profiles (Line M) at a load level of
114 kN (26 kip) per jack ....................................................................... 351
Figure 6.63 Comparison of center deflection profiles (Line M) at 1st observed
cracking loads ........................................................................................ 351
Figure 6.64 Comparison of center deflection profiles at 2nd observed cracking
loads (Line M) ....................................................................................... 352
Figure 6.65 Detail of cracking observed at ultimate capacity- top view of asbuilt deck ............................................................................................... 353
Figure 6.66 Detail of cracking observed at ultimate capacity- top view of FRP
rehabilitated deck................................................................................... 353
Figure 6.67 Top view of deck slab tested to ultimate capacity after removal of
loose concrete ........................................................................................ 354
xxi
LIST OF TABLES
Table 2.1 Representative bridge systems constructed using FRP structural SIP
formwork for bridge decks ...................................................................... 31
Table 3.1 Performance characteristics of the Grade 60 reinforcing steel .................... 50
Table 3.2 Performance characteristics of the concrete ................................................ 51
Table 3.3 Lamina information and nomenclature [120] ............................................... 55
Table 3.4 Experimental mechanical properties of bottom plate (Mark 4) [119] .......... 56
Table 3.5 Mechanical properties of deck panel components [120] ............................. 57
Table 3.6 Summary of instrumentation used for test specimen #1 ............................. 65
Table 3.7 Resolution attainable by linear potentiometers used on test specimen ....... 67
Table 3.8 Static loading protocol................................................................................. 83
Table 3.9 Effective stiffness ratios of FRP side slab from linear potentiometer
data .......................................................................................................... 90
Table 3.10 Effective stiffness ratios of RC side slab from linear potentiometer
data .......................................................................................................... 90
Table 3.11 Comparison of effective stiffness ratios found using two different
methods of calculating effective stiffness ............................................... 93
Table 3.12 Effective stiffness ratio of FRP side slab from linear potentiometer
data ........................................................................................................ 135
Table 3.13 Effective stiffness ratio of RC side slab from linear potentiometer
data ........................................................................................................ 135
Table 3.14 Residual deflections of deck slabs (from central linear
potentiometers) ...................................................................................... 137
Table 3.15 Safety factors for different limit states in terms of permit loadings ........ 147
Table 3.16 Cyclic loading protocol for test specimen #1 .......................................... 148
Table 3.17 Center deflections during cyclic loading phase I (84 kN target load) ...... 150
Table 3.18 Center deflections during cyclic loading phase II (168 kN target
load) ....................................................................................................... 150
Table 3.19 Effective stiffness ratios for FRP panel side during cyclic loading
phase I .................................................................................................... 151
Table 3.20 Effective stiffness ratios for RC side during cyclic loading phase I ....... 151
Table 3.21 Effective stiffness ratios for FRP panel side during cyclic loading
phase II .................................................................................................. 153
Table 3.22 Effective stiffness ratios for RC side during cyclic loading phase II ...... 153
Table 3.23 Strain progresssion for composite side gages under target loadslocated directly below the composite side actuator ............................... 155
Table 3.24 Strain progresssion for reinforced concrete side gages under target
loadslocated directly below the RC side actuator ............................... 156
Table 3.25 Strain progresssion for transverse composite side gages under target
loadslocated along the centerline, near the stems ............................... 158
Table 3.26 Separation progresssion for linear potentiometers located along the
test speciment centerline, near the stems ............................................... 159
Table 4.1 Properties of Sikadur 55 SLV resin system [156] ..................................... 166
Table 4.2 Static loading protocol for testing of repaired specimen ........................... 168
xxii
Table 4.3 Effective stiffness ratios of original FRP panel side slab .......................... 175
Table 4.4 Effective stiffness ratios of original FRP panel side slab, end of cyclic
phase ...................................................................................................... 176
Table 4.5 Effective stiffness ratios of repaired FRP panel side slab ......................... 176
Table 4.6 Strain values for CM3L at different load levels ........................................ 192
Table 4.7 Strain values for CM3T at different load levels ........................................ 194
Table 4.8 Strain values for DTL8-3 at different load levels ...................................... 196
Table 4.9 Strain values for DTT10-3 at different load levels .................................... 197
Table 4.10 Cyclic loading protocol for repaired specimen ....................................... 204
Table 4.11 Effective stiffness ratios for FRP panel side of original test specimen
during the 1st 500 cycles of cyclic loading phase II (2x AASHTO
service load)........................................................................................... 207
Table 4.12 Effective stiffness ratios for FRP panel side of repaired test
specimen during cyclic loading (2x AASHTO service load) ................ 207
Table 4.13 Strain progresssion for composite side gages under target loadslocated directly below the composite side actuator ............................... 209
Table 4.14 Separation progression for linear potentiometers located along the
test speciment centerline, near the stems ............................................... 211
Table 5.1 Comparison of 1st and 2nd test specimens .................................................. 216
Table 5.2 Properties of Simpson Epoxy-Tie adhesive system [157]......................... 222
Table 5.3 Summary of instrumentation used for test specimen #1 ........................... 227
Table 5.4 Strain gages used in test specimen #2 ....................................................... 230
Table 5.5 Static loading protocol for test specimen #2 ............................................. 242
Table 5.6 Effective stiffness ratio of FRP side slab from linear potentiometer
data ........................................................................................................ 246
Table 5.7 Effective stiffness ratio of RC side slab from linear potentiometer
data ........................................................................................................ 246
Table 5.8 Residual deflections of deck slabs (from central linear
potentiometers) ...................................................................................... 265
Table 5.9 Cyclic loading protocol for test specimen #2 ............................................ 270
Table 5.10 Effective stiffness ratios of the two slabs throughout cyclic loading ...... 274
Table 6.1 Loading protocol for as-built test specimen .............................................. 302
Table 6.2 Physical properties of pultruded CFRP strengthening product options
[162-164] ............................................................................................... 318
Table 6.3 Different NSM FRP strengthening options ............................................... 321
Table 6.4 Tensile properties of SikaDur 30 resin system [169] ................................ 323
Table 6.5 Tensile properties of SIKA CarboDur S512 CFRP strips [169] ............... 323
Table 6.6 Tg progression of CarboDur 30 resin used in NSMR installation ............. 326
Table 6.7 Loading protocol used for FRP rehabilitated specimen ............................ 333
Table 7.1 Performance comparison between FRP panel system and RC system ...... 358
xxiii
ACKNOWLEDGEMENTS
The author extends her appreciation and admiration to her dissertation advisor,
Professor Vistasp Karbhari, for his uncompromising dedication to his students and his
passionate enthusiasm for his work. Without his time and efforts, this work would not
have been possible.
Partial funding for this research was provided by the California Department of
Transportation; Dr. Charles Sikorsky was the fund manager.
Sika Corporation
donated composite materials and Sika representative, Adriano Bortolin, provided

technical assistance for the resin injection repair and near surface mounted composite
rehabilitation stages of this research. The author offers appreciation to her dissertation
committee members, Structural Engineering Professors Gilbert Hegemier, Francesco
Lanza di Scalea, and Hyonny Kim, as well as Mechanical Engineering Professors
Vlado Lubarda and Prab Bandaru, for their insightful feedback on this research.
Technical support from the staff at the Powell Structures Laboratory at the
University of California, San Deigo, particularly the assistance of Dr. Christopher
Latham, Andrew Gunthardt, Bob Parks, Charles Stearns and Edward Stovin is greatly
appreciated. Fellow graduate students within Dr. Karbharis research group and within
the structural engineering department, particularly Don Philippi, Griffin Brungraber
and SeungDae Kim provided valuable assistance with experiments as well as many
insightful conversations.
Lastly, the author wishes to thank her husband and family who have been
tremendously supportive during this process.
xxiv
VITA
2001-2004
Assistant Research Scientist, La Jolla Bioengineering Institute
2004-2006
Research Assistant, Department of Structural Engineering

2006
B.S., Structural Engineering

2006-2009
Research Assistant, Department of Structural Engineering

2007-2008
Teaching Assistant, Department of Structural Engineering

2008
M.S., Structural Engineering

2009
Instructor, Department of Structural Engineering

2009
Ph.D., Structural Engineering

xxv
ABSTRACT OF THE DISSERTATION
ABSTRACT
Development of Composite Renewal Systems for Rapid
Rehabilitation and Construction of Bridge Decks
by
Anna Beth Pridmore
Doctor of Philosophy in Structural Engineering

University of California, San Diego, 2009
Professor Vistasp M. Karbhari, Chair

Professor Gilbert A. Hegemier, Co-Chair
The deterioration of steel in aging reinforced concrete bridges is a continual

problem which could benefit from improved rehabilitation techniques that take
advantage of enhanced and more durable materials such as fiber reinforced polymer
(FRP) composites. Appropriately designed hybrid material systems benefit from the
performance and durability advantages of FRP materials yet remain more cost
effective than comparable all-composite systems.
Development of rapid rehabilitation systems for the decks of concrete box
girder bridges, which are increasingly common throughout the United States, is
presented.
One goal of this research is to assess and validate the use of FRP
xxvi
composite panels for use as both stay-in-place formwork and as the bottom
longitudinal and transverse reinforcement in the deck of concrete box girder bridges.
Performance assessments for full-scale two-cell box girder bridge specimens through
monotonic and extensive cyclic loading provided validation for the FRP panel system
bridge deck as a viable rehabilitation solution for box girder bridge decks. The FRP
panel system performed comparably to a conventionally reinforced concrete bridge
deck in terms of serviceability, deflection profiles, and system level structural
interaction and performed superior to the RC bridge deck in terms of residual
deflections, and structural response under cyclic loading. Assessment of a damaged
FRP panel bridge deck system, which was repaired using a resin injection technique,
showed superior performance for the repaired system in terms of integrity of the FRP
panel interface and cyclic response.
Rapid rehabilitation techniques for strengthening reinforced concrete box
girder bridge deck overhangs using near-surface-mounted (NSM) carbon fiber
reinforced polymer (CFRP) were also evaluated. Analytical predictions of load
carrying capacity and deflections provided correlation with experimental results, and
the developed analysis methods provide an effective design tool for future research.
Results from the laboratory testing of a bridge deck overhang strengthened with FRP
showed significant increases in load carrying capacity as well as deformation capacity
as compared to the as-built specimen without FRP. This research provides enhanced
understanding of hybrid structures and indicates significant potential for rehabilitation
applications to concrete box girder bridges.
xxvii
1
1.1
INTRODUCTION
BACKGROUND AND MOTIVATION FOR RESEARCH

Usually designed for a fifty year life span, the average bridge in the United
States is now 43 years old [1]. In 2008, the American Association of State Highway
and Transportation Officials (AASHTO) estimated that approximately $140 billion
would be necessary to repair every currently deficient bridge in the country [2]. The
cost to continually eliminate all bridge deficiencies that would occur over the next
fifty years was estimated to be $850 billion, however annual government investment
in rehabilitation and replacement of bridges falls more than 40% short of the annual
investment necessary to achieve this goal [3]. This necessary financial investment
creates significant motivation for development of more effective bridge rehabilitation
techniques.
The element within a bridge system that usually requires the most maintenance
is the bridge deck, which typically lasts half as long as the rest of the bridge, 35 years
versus 68 years respectively [4]. Because the bridge deck is directly impacted by
traffic loading, it is most susceptible to cyclic loading induced damage and
environmental degradation [5]. Nearly 70% of the existing bridge decks in the United
States are cast-in-place reinforced concrete [6]. Cracks that form in the concrete allow
moisture and deicing salts to penetrate into the deck and corrode the reinforcing steel
rebar [7]. The corrosion of steel reinforcement is a significant problem that adversely
affects the long term durability of the concrete because the steel rebar expands as it
2
rusts, causing the surrounding concrete to crack and spall, and further accelerate the
deterioration of the bridge deck [8].
Deterioration in steel reinforced concrete bridge decks can be reduced by
improving design practices, such as ensuring proper cover for the steel reinforcement,
and through the use of higher quality conventional materials. Some techniques for
reducing the rate of deterioration include the use of epoxy coated or stainless steel
rebar, the application of surface treatments with cathodic protection and sealants to
inhibit penetration of salts into the concrete slab, the filling of cracks using a grout or a
low viscosity resin, and the use of a wearing surface [9]. These construction methods
are typically costly and offer only a temporary barrier to preventing penetration of
salts in to the concrete slab. For bridge decks with extensive deterioration, a common
technique is the removal of the old deck and construction of a new cast-in-place
concrete deck [10]. This costly solution does not address the inherent problems
associated with corrosion prone steel reinforcement.
1.2
FRP REHABILITATION OF CONCRETE BRIDGE DECKS

The use of FRP composites is becoming an increasingly popular alternative
due to the many performance and potentially enhanced durability benefits of FRPs
such as superior resistance to damage from freeze-thaw cycles and deicing chemicals
[5, 11]. FRP composites offer high specific strength and stiffness, enhanced fatigue
life and corrosion resistance, and highly anisotropic properties that allow for improved
tailorability to numerous design applications [5]. While FRP material costs for FRP
systems are relatively expensive compared to traditional construction materials such as
3
concrete and steel, FRP systems provide opportunities for reduced labor and
equipment costs as well as accelerated construction schedules [12].
FRPs have been applied to bridge deck applications in the form of allcomposite systems [13-28] or as a combination of FRPs with conventional materials
such as concrete, known as a hybrid system [29-35].
Further information on
applications of FRP composites to bridges can be found in recent reviews presented by

Cheng and Karbhari [36], Bank [37], Bakis et al. [38], and Keller [39]. While all FRP
decks are an attractive option, their cost is still relatively high [40, 41]. A major
advantage of hybrid FRP-concrete systems, which include FRP rebar [31, 42-54], grid
reinforcement [40, 55-64], and stay-in place (SIP) structural formwork [40, 58, 59, 6573], is their efficient use of materials.
By designing a system where the high
performance FRP composites, which are best used in tension, are utilized in tension
regions and the economical concrete, which is good in compression, is utilized in
compression regions, material usage can be optimized.
This judicious material
selection creates a more cost-effective system than an all-composite design, while

taking full advantage of the excellent tensile strength, tailorability and potential for
corrosion resistance provided by FRP materials [5]. Note that FRP rebar generally has
a lower modulus than steel and therefore will need to be used either at a great volume
fraction than steel, or over shorter spans [38]. Similar constraints exist for grids,
although in both cases the potential durability advantages are important for niche
applications. Structural stay-in-place FRP formwork however serves as an example of
an optimal hybrid the composite is used as the formwork as well as the bottom layer
of tensile reinforcement for the bridge deck, and the concrete is used in compression
4
both systems in their best modes of structural response. In addition to potential
durability benefits, SIP structural forms can offer economic advantages by reducing
labor costs through elimination of the formwork and falsework necessary for
construction of a cast-in-place concrete bridge deck. The use of FRP structural panels
as both the formwork and tensile reinforcement in the positive moment regions of a
reinforced concrete box girder bridge deck is a rehabilitation system that was assessed
and validated as part of this research.
1.3
REHABILITATION NEEDS FOR REINFORCED CONCRETE BOX GIRDER BRIDGES

The current research focuses on addressing rehabilitation needs for reinforced
concrete box girder bridges. Reinforced concrete box girder bridges are commonly
used throughout the United States and comprise 45% of Californias bridge inventory,
with the majority of these bridges built using cast-in-place concrete [74]. The use of
box girder bridges is increasingly common in part due to advantages provided by
improved resistance to torsion and the capability for long span lengths. Problems
specific to the box girder bridges are that they are more difficult to maintain because
of need for access to a confined space and that deck replacement is difficult due to
issues of instability in the box.
The first rehabilitation option addressed in this research involves the use of
FRP structural formwork panels in conjunction with regular strength concrete to
provide an effective design alternative to traditional reinforcement systems for a
reinforced concrete box girder bridge. This system can be used as either a method of
repair for bridge decks with severe or extensive damage or used for new bridge deck
5
construction. The design of the system provides significant reserve strengths for the
deck panel system and constrains damage within the concrete to a more localized
region between stiffeners, thereby allowing damaged concrete to be removed and new
concrete to be cast without requiring access to the inside of the box. A resin injection
repair technique of the developed FRP panel system was also examined as a further
means to accommodate effective rehabilitation of box girders without requiring access
to the inside of the box.
Another rehabilitation method examined in this research addresses a problem
associated with older box girder bridges, which lack sufficient strength to
accommodate the increased dead loads due to the addition of sound walls and traffic
barriers onto the overhangs of the bridge decks.
The conventional method of
increasing the capacity of the overhang region is to remove the overhang and portions
of the adjacent bridge deck and stem wall, and then rebuild the section with adequate
reinforcement using conventional reinforced concrete. The current research method
addresses the use of near surface mounted CFRP strips for flexural strengthening the
overhang region of the bridge deck. This method allows the existing overhang to
remain intact and significantly reduces construction time over a partial deck rebuild.
6
1.4
RESEARCH GOALS AND OBJECTIVES

The primary goal for this research is to evaluate rapid rehabilitation methods
that employ FRP materials for use in the decks of concrete box girder bridges. The
major objectives for achieving this goal include:
1. Assess and validate the use of stay-in-place FRP composite formwork
panels for use in the decks of box girder bridges;
2. Characterize the system level behavior of the FRP panel bridge deck
system;
3. Provide fundamental performance comparisons between the FRP panel
bridge deck system and a conventional reinforced concrete deck system;
4. Determine limit states and develop preliminary design guidelines for the
FRP panel bridge deck system;
5. Evaluate the ease and effectiveness of techniques to repair the FRP panel
bridge deck system if damage to the bridge deck were to occur;
6. Examine additional rapid rehabilitation techniques such as the use of near
surface mounted (NSM) FRP reinforcement for strengthening the negative
moment regions of box girder bridge deck overhangs; and
7. Develop analytical performance predictions and design recommendations
for rehabilitation of bridge decks using NSM FRP reinforcement.
7
1.5
SCOPE OF PROJECT
This research is organized into seven chapters.
Chapter 1 introduces the
problems addressed by this research, the research goals and objectives and the scope
and outline for the research. Chapter 2 presents a literature review for the major topics
covered in this research.
Chapter 3 presents the development, experimental testing program, and results
from an investigation of a full scale two -cell box girder bridge which incorporates the
developed hybrid FRP-concrete bridge system, where prefabricated composite panels
act as both the formwork and reinforcement for the bridge deck. Chapter 4 addresses
the implementation of a epoxy injection technique used to repair the damage caused to
the concrete-composite interface of the post-failure FRP panel system deck system.
The details of the repair technique and analysis of the test results are presented.
Chapter 5 presents the details and test results from an experimental investigation
which focuses on the structural response of the FRP panel system under cyclic
loading.
Chapter 6 presents the experimental and analytical investigation of the
strengthening of a reinforced concrete bridge deck overhang with near surface
mounted composite strips. Conclusions and recommendations for future work related
to this research are discussed in the final chapter.
2
2.1
LITERATURE REVIEW
OVERVIEW
The first section of this literature review provides information relating to
formwork for construction of slab-on-girder bridge decks. Different materials and

designs used as formwork for concrete slab-on-girder bridge decks are presented and
the relative merits of the different systems are addressed. Particular focus is given to
the progression of research relating to FRP stay-in-place structural formwork systems
for bridge decks in order to provide context for the current research.
An overview of typical repair methods for reinforced concrete bridge decks
given in Section 2.3 offers a background for the epoxy injection repair technique
implemented to reinstate load transfer between the FRP panels and concrete on a
failed test specimen, which is discussed in Chapter 5.
Next, research related to the use of near surface mounted composites for
strengthening concrete structures is provided in Section 2.4. This application of near
surface mounted composites provides rapid rehabilitation for bridge decks and often
allows repair where conventional mechanisms are too costly or not possible.
comparison between surface mounted and near surface mounted FRP as well as an
overview of the different failure modes of NSM, and available specifications are also
presented.
2.2
FORMWORK FOR CONSTRUCTION OF SLAB-ON-GIRDER BRIDGE DECKS

The primary purpose of deck formwork is to provide a shape for the concrete
deck and to support the dead load from the reinforcement and fresh concrete as well as
the live load from construction materials, workers and equipment. The three major
objectives considered when designing formwork systems are quality, safety and
economy [75].
Quality forms must provide sufficient stiffness to maintain desired
tolerances for the cast concrete and must safely support all dead and live loads without
system collapse or safety issues. Formwork costs may range from 35 to 60% of the
total cost of the concrete structure; therefore, economy of formwork design can
provide a significant impact on the overall construction budget for a project [76, 77].
In addition to direct labor costs, the speed of construction is often controlled by
formwork operations. In order for work to proceed smoothly and cost effectively, it is
necessary to carefully select the bridge deck formwork system to best meet a
construction projects job requirements.
2.2.1
Temporary Formwork versus Stay-in-Place Formwork

Removable formwork supports bridge deck construction loads and freshly
placed concrete until the deck concrete cures sufficiently to allow removal of the
forms. If the deck concrete has not achieved sufficient strength to fully support it own
weight without excessive deflections, the deck is reshored with temporary vertical
supports until full design strength for the deck is obtained [78]. After the stripped
10
forms are inspected and any necessary repairs have been completed, the forms may be
reused.
Temporary formwork systems are often constructed of wood supports and
plywood sheathing, with designs similar to the layout shown in Figure 2.1 [76]. This
type of form is usually constructed in place on the job and reuse of form material is
limited to the salvaging of individual boards or timbers [75].
Because this type of
system is very labor intensive and requires a high-quality labor force, prefabricated
reusable forms which can be constructed more quickly with less on-site preparation
have become more prevalent. Reusable forms for bridge decks are most commonly
made of wood, steel or aluminum, or a combination of these materials, which can be
custom fabricated for a specific project or can be part of a proprietary system.
Figure 2.1 Conventional wooden shored formwork layout for deck (after [76])
Formwork systems can also be permanently left in place for the life of the
bridge deck. Stay-in-place (SIP) bridge deck forms (also known as lost or permanent
11
forms) are usually used on short spans where intermediate supports are not required
and are made of glass fiber reinforced cement, fiber reinforced polymers, corrugated
steel sheets, or precast concrete panels [75].
While temporary formwork systems
can reduce material costs by allowing reuse of the forms, potential benefits from the
use of stay-in-place forms include simplified construction, reduction in construction
time and labor costs, and enhanced safety. By eliminating the need for falsework to
support the deck forms and reshoring after forms are removed, construction times and
labor costs can be reduced.
Safety during the construction process is enhanced,
because the stripping of formwork, which is often located in difficult and confined
spaces that leave workers vulnerable to falling tools or other objects, is no longer
required [79]. The SIP forms also provide a safe working platform at an early stage in
the deck construction process, which is a major advantage when the bridge is spanning
a river, active roadway, or other hazardous location [80]. The modularity of stay-inplace forms provides the potential for increased standardization and the simple
connection details commonly associated with SIP forms can simplify the construction
process [69]. SIP forms can also act as structural elements for the bridge deck and
help to support service loads, which can allow for reductions in the amount of required
steel reinforcement within the deck. However, depending on the type of stay-in-place
form used, disadvantages can include trapping of moisture that collects on top of the
SIP forms as well as damage and corrosion problems (especially associated with steel
SIP forms) due to the trapped moisture [68]. In addition, forms which have smooth
surfaces prevent composite action from developing with the concrete deck.
12
2.2.2
Stay-in-Place Formwork Made from Conventional Construction Materials

Use of precast concrete panel forms has been documented for use in building
and bridge construction since before the mid 1930s [75]. Concrete forms can either be
plain, reinforced, or prestressed, and can be cast at the job site or in a factory [81].
The forms can act solely as formwork or they can also act compositely with the castin-place concrete and become part of the load bearing structure. An early example of
a typical structurally participating forms made of precast concrete panels is a Northern
Illinois Toll Highway bridge, built in 1956, which uses precast prestressed slabs that
spanned between prestressed girders. The panels had a roughened top surface as well
as hairpin dowels to improve the bond between the panels and the cast-in-place
concrete such that composite action could be ensured. The SIP panel forms provided
some of the positive moment deck reinforcement and added to the strength of the full
deck through composite action.
Use of precast prestressed concrete SIP forms as part of a concrete bridge deck
system without any internal reinforcement, has also been reported [82]. This system is
a continuation of the original concept that was developed by Mufti et al., which relied
on internal arching action of a steel-free cast-in-place concrete or precast concrete
deck and used steel straps to provide transverse confinement for the deck [83-85]. The
Salmon River Bridge in Nova Scotia, Canada was the first of many steel free bridges
to rely on arching action [86], and provisions for designing steel free slabs that rely on
arching action have been incorporated into the Canadian Highway Bridge Design
Code (CHBDC) [87]. A grid of nominal reinforcement, either made of steel of GFRP
is recommended to control unsightly, but not structurally significant, cracks that had
13
developed on the bottom of the deck slabs [88]. Use of light gage-steel formwork that
takes the place of the steel straps to serve as the tension chord has also been presented
and design modifications to the CHBDC code were provided by Bakt et al [89].
While the use of corrugated metal SIP forms for bridge decks is prevalent,
significant concerns are associated with the difficulty in achieving composite action
due to the smooth surface of the forms as well as corrosion problems and freeze/thaw
damage of the SIP metal forms due to trapped moisture and chlorides [68]. Therefore,
thirteen US states do not allow the use of metal SIP forms due to concerns that the
presence of the metal SIP forms may accelerated damage of the bridge deck and cause
higher contents of moisture and chlorides to remain within the deck [90]. In addition,
AASHTO does not permit concrete decks with metal SIP formwork to be considered
as composite slabs for design purposes [91], even though composite slabs are routinely
used in building design.
Recent developments of glass fiber reinforced cement SIP forms for bridge
decks were presented by Kim et al. [92, 93] and Bank et al. [94]. Bank et al. also
provided a model design and construction specification on fiber reinforced nonstructural SIP formwork panels, which span narrow gaps of less than 1 m (39.4 in), for
use in construction of concrete highway bridge decks [66].
2.2.3
FRP Stay-in-Place Structural Formwork

In the past 20 years, research interest has lead to advances in the development
of hybrid systems that combine FRP as stay-in-place structural formwork with

concrete. Use of a hybrid system, which involves cast in place concrete, together with
14
stay-in-place FRP structural formwork allows for the development of a more cost
effective system. By appropriate design, that uses concrete which is good in
compression and is less expensive than FRP material, in conjunction with FRP
material, which are good in tension, materials usage can be optimized. The use of FRP
as permanent formwork can also facilitate improved construction schedules, and has
the potential for improved bridge deck durability by reducing or eliminating the need
for corrosion-prone steel reinforcement.
In 1989, the concept of a pultruded GFRP membrane as the structural
formwork for a lightweight concrete bridge deck was proposed by Bakeri [95]. The
GFRP membrane design shown in Figure 2.2 was selected from ten different hybrid
GFRP-concrete conceptual designs based on ease and cost of manufacturing, as well
as analytical calculations predicting mechanical performance of the system.
Alternative conceptual designs included light weight concrete filled filament wound
GFRP domes, pultruded GFRP vaults, and filament wound doubly-curved membranes
as well as pultruded GFRP trusses with light weight concrete overlays. Mechanical
performance of the system was analyzed based on quasi-isotropic layups, however no
experimental data was available.
Figure 2.2 Concrete filled membrane deck system (after [95])
15
2.2.3.1 Composite Slabs Using FRP Panels with T-upstands

In 1990, the concept of a composite slab system constructed using pultruded
FRP SIP formwork with concrete fill was proposed by Hillman and Murray as a
lightweight floor system for steel framed buildings [96]. The pultruded deck consisted
of 76.2 mm (3 in) deep inverted T-beams, which provided the stiffness for the system,
that were connected together by plates located approximately 25.4 mm (1 in) from the
top flanges. The plates acted as the formwork for a 25.4 mm (1 in) concrete slab
placed on top of the flanges, as shown in Figure 2.3. When full composite action
between the FRP SIP formwork and the concrete was assumed, Hillman and Murray
suggested that a 50-60% weight reduction as compared to the cold-formed steel
decking reference slabs could be achieved.
Figure 2.3 Composite slab with structural formwork (after [96])
A feasibility study of the concept proposed by Hillman and Murray was

conducted by Hall and Mottram using commercially available pultruded GFRP panels
that consisted of two T-upstands with a continuous base [73]. Twelve beams that each
16
were 120 mm (5 in) wide and had a span of 2 meters (6 ft 10in) were tested to failure
using a four point bend test. The beams used different configurations of the pultruded
FRP panels and different concrete heights, as shown in Figure 2.4. The system
behaved in linear elastic fashion within the serviceability range. However, significant
horizontal shear slippage was observed between the concrete and the FRP sections at
loads significantly above the serviceability range, which severely reduced the ultimate
capacity of the member. In order to enhance the concrete-composite interface, the
surfaces of the pultruded panels which came into contact with concrete were coated
with an adhesive and the concrete was cast before the adhesive had cured. The beams
with adhesive treated FRP-concrete interfaces displayed full composite action and
showed significantly improved performance over the beams without adhesive
treatment. For the beams which did not fail due to FRP-concrete debonding, the
observed failure mode was diagonal tension shear cracking in the concrete. The
observed failure mode was attributed to the beams design, which was over-reinforced
in flexure and to the absence of shear reinforcement within the concrete.
Figure 2.4 Beam specimens with FRP SIP forms (after [73])
Keller et al. reported on a further development of FRP panels with T-upstands

for use as SIP structural formwork [97-99]. In order to reduce the self weight of the
17
systems, a sandwich panel was developed that consisted of a pultruded FRP sheet with
T-upstands for the tension skin and structural formwork, a lightweight concrete (LC)
for the core, and a normal weight concrete (NC) on top for the compression skin, as
shown in Figure 2.5 (a). Eight beams, each 400 mm (16 in) wide, 200 mm (8 in) deep
and 3.6 meters (12 ft 2 in) long, were tested in flexure using a three-point bending
setup with two types of LC and two types of FRP/LC interface: an adhesive bond
versus only mechanical interlock between the LC and T-upstands.
(a) Cross section of sandwich panel

NC = Normal weight concrete
LC = Light-weight concrete
(b) Close-up of failure surface

at concrete-FRP interface
Figure 2.5 Sandwich panel with a lightweight concrete core, concrete in compression
zone, and FRP panel with T-upstands in tension zone [97, 98]
The beams that relied on mechanical interlock between the LC and FRP panels
exhibited significant slippage between the core and the panels, as shown in Figure 2.5
(b), thereby reducing composite action for the section. The adhesive bond between the
18
LC core and FRP panels provided sufficient interface strength such that full composite
action was observed prior to the observed failure mode of horizontal shear within the
LC core.
Successful serviceability performance and overall results showed the
feasibility of the proposed hybrid deck system.

In 2006, Bank et al. reported on the experimental testing of commercially
available pultruded FRP planks used as formwork and reinforcement for a concrete
member [65, 71]. The FRP planks with T-upstands were coated with different types
of aggregate that were bonded with epoxy to the FRP planks to increase mechanical
interlock and prevent slip between the concrete and the FRP panels. The feasibility of
this system was experimentally examined using a total of seven 203 mm wide (8 in)
by 178 mm (7 in) deep beams with spans of 1.1 m (3 ft 7 in) and 1.8 meters (6 ft)
tested in three-point bending While partial debonding between the FRP plank and the
concrete was observed for all the beams, no slip between the concrete and aggregate
coated FRP planks was noted. In contrast, the control specimen, which consisted of
the same type of FRP plank without any surface treatment, exhibited significant slip
and failed at approximately half the failure load exhibited by comparable beams with
aggregate coated FRP planks. The aggregate coated FRP plank system compared
favorable to a steel reinforced system indicating the systems potential for serving the
dual functions of formwork and tensile reinforcement.
Based on the results of this study, the examined FRP plank SIP formwork and
reinforcement system was used in the construction of a new bridge in Wisconsin
located over Coffee Creek at Black River Falls, as shown in Figure 2.6. This two lane
30 m span bridge was completed in 2008 consists of a reinforcement free concrete
19
deck (except for guardrail attachments) placed on precast concrete bulb-Tee girders in
which aggregate-coated FRP planks are utilized as SIP formwork and secondary
reinforcement for the bridge deck.
The same research group also investigated a FRP SIP structural formwork
system consisting of a pultruded GFRP plate which is bonded to a GFRP grid formed
by I bars that are intersected by periodically spaced GFRP rebar which have been
inserted through holes drilled into the I bars [69]. Multiple one span and two span
slabs were tested under flexural as well as combined flexural and axial loading. ACI
440 design guidelines were found to accurately predict flexural and flexural-shear
capacities and additional capacity prediction techniques were proposed for flexural
and flexural-shear under combined loading.
Figure 2.6 Use of commercially available FRP plank as SIP forwork and secondary
reinforcement in reinforcement-free deck [66]
20
2.2.3.2 FRP Box Beams with Concrete in the Compression Zone

In the early 1990s, Deskovic et al. tested the short term and long term
behavior of GFRP box beams with concrete in the compression zone and CFRP in the
tension zone [100-102]. This work built upon previous studies of GFRP box beams
from the 1980s [103, 104], which established the following design deficiencies:
1 . failure of the GFRP box beam is usually catastrophic, because composite
materials are observed to behave in a linear elastic fashion until failure,
without significant plastic deformation as seen in steel or reinforced
concrete;
2. the compressive flange is significantly weaker than the tensile flange, due
to local buckling phenomena and because a unidirectional of GFRP has a
compressive strength of approximately half its tensile strength; and
3. design is usually governed by stiffness, due to the relatively low stiffness
of GFRP, which results in a overly costly design due to the required use of
additional composite material in order to satisfy stiffness requirements.
In order to mitigate the aforementioned problems, concrete was chosen for use
in the compression zone of the section because of concretes excellent compressive
strength and stiffness to cost ratio. The use of concrete in the compression zone
allowed for a reduction in the amount of composite material needed, thereby providing
significant reductions in material costs for the system. To minimize fabrication costs,
the upper flange of the GFRP cross section acted as formwork for casting the concrete
as shown in Figure 2.7.
A thin layer of unidirectional CFRP was bonded to the
sections tensile flange in order to enhance the sections rigidity because CFRP
21
possesses a higher stiffness than GFRP. In addition, the lower strain-to-failure of
CFRP as compared GFRP provided a warning mechanism prior to ultimate collapse of
the system (pseudo-ductility).
Concrete
GFRP
CFRP
Figure 2.7 FRP box beam with concrete in compression zone (after [100])
A total of seven full sized filament wound GFRP box beams of the type shown
in Figure 2.7 with dimensions of 180 mm (7.3 in) by 300 mm (12.2 in) and a span
length 3 meters (10 ft) were tested in three point bending, with three beams tested for
short term performance and the other four beams tested to establish long term
characteristics. The first beam, which had no shear connectors between the concrete
and the GFRP beam, failed due to fracture of the CFRP followed by GFRP-concrete
interface debonding.
An examination of the fractured surfaces suggested the lower
bond strength was partially attributed to poor surface preparation in which silane, a
component used during the composite manufacturing process, was not properly
removed from the surface of the composite.
Subsequent beams that used shear
connectors to increase the bond strength showed the anticipated pseudo-ductile failure
22
behavior of CFRP fracture followed by concrete crushing at a higher load [100].
Results of creep and fatigue testing also provided good agreement with analytical
predictions.
Further research related to FRP box beams with a thin concrete top layer was
reported by Canning et al. [105] and Hulatt et al. [106].
Canning et al. examined the
effectiveness of different techniques used to develop shear transfer between the top
layer of concrete and the FRP box beam. Experimental results from flexural testing of
six beams with 1.5 m (5 ft) spans indicated that the use of an adhesive applied directly
prior to concrete casting was the most practical of the techniques outlined that still
achieved a high level of composite action. Hulatt et al. experimentally tested nine
vacuum bagged GFRP box beams with concrete in the compression zone and CFRP in
the tension zone under both static and fatigue four-point bending. The beams were
stiffened with composite diaphragms adhesively bonded to the inside of the beam as
shown in Figure 2.8 in order to prevent buckling of the thin webs.
Figure 2.8 FRP box beam with concrete compression flange (after [106])
23
The beams behaved compositely throughout testing and the primary observed failure
mode was concrete crushing. The results showed appreciable resistance to sustained
and fatigue loading as well as close agreement between experimental and theoretical
work [106].
A further development of the concept proposed by Descovic et al. was
reported by Kitane et al in 2004 [29, 30, 107], which involves three hollow trapezoidal
GFRP tubes bonded together and partially filled with concrete in the compression
zone, as shown in Figure 2.9. CFRP strips were also bonded to the bottom side of the
trapezoidal tubes in order to increase the stiffness of the system. Specimens with
spans ranging from 2.31 m (8 ft) to 4.95 m (16 ft 10 in) were tested under multiple
loading scenarios including three-point bending, negative bending, eccentric positive
flexure and shear dominated loading.
Figure 2.9 Hollow trapezoidal tubes partially filled with concrete (after [108])
The stable structural response of the system under different loading cases
showed the feasibility of the system and matched closely with analytical predictions.
The researchers had proposed a 6.5 m (22 ft 1 in) long, 4.06 m (13 ft 10 in) wide
prototype bridge system which was designed as a simply supported steel girder bridge
24
with the examined hybrid FRP-concrete deck system.
In 2008, Alnahhal et al.
reported on the structural response of a 3/4 scale version of the FRP-concrete deck on
steel girders specimen, which was tested under seven different. loading conditions
[109]. The system satisfied the AASHTO live load deflection limits and the hybrid
deck and steel girders were observed exhibit a partially composite action under service
loading conditions.
In 2008, Honickman experimentally and analytically examined the
performance of GFRP box sections made of trapezoidal GFRP plates connected to flat
GFRP sheets with concrete poured on top [79]. A total of eight 3.10 m (10 ft 5 in)
long specimens were tested in four point bending under either positive or negative
flexure and the results were compared to conventional reinforced concrete slabs. The
three variables for the treatment of the cavity between the GFRP trapezoidal plate and
top GFRP flat sheet were partially filled with concrete such that a void was cast in the
tension region of the concrete, completely filled with concrete, or not filled with any
concrete as shown in Figure 2.10. All specimens had a thin steel mesh and a 610 mm
thick concrete layer cast on top of the concrete upper plate as well as concrete
diaphragms in the end regions to allow for loading in negative bending.
Figure 2.10 Corrugated GFRP SIP formwork acting as reinforcement for a concrete
beam (after [79])
25
The all-GFRP box section with the concrete slab showed the best performance
with a significantly higher strength as compared to any other specimen and a reduction
in weight of 50% and 65% relative to the partially concrete filled and completely filled
specimens, respectively. The flexural strength for the specimens loaded in positive
bending was 45% to 90% higher than in negative bending, depending on the crosssectional configuration. Different bond mechanisms were also examined and the wet
adhesive bonding of fresh concrete to the GFRP plate was preferred in terms of
structural performance as well as ease and speed of fabrication, however specimens
which incorporated headed studs in conjunction with adhesive bonding were able to
achieve a mode of flexural failure with concrete crushing rather than the debonding
failure observed in the other specimens.
2.2.3.3 FRP Panels with Rectangular Stiffeners

In 2000, pultruded GFRP panels were installed as SIP structural formwork for
one and a half spans of the five span Salem Avenue bridge deck replacement project in
Dayton, Ohio [110-113]. The three lane, 207 m (704 ft) long, bridge used four
different FRP deck systems which were supported on steel girders spaced at 2.64 m (8
ft 8 in) on center. One of the deck systems, shown in Figure 2.11, consisted of a 216
mm (8.5 in) thick layer of concrete cast on top of a mat of GFRP rebar for the top
layer of reinforcement and pultruded planks with 76 mm (3 in) high tubular stiffeners
as the formwork and lower level of reinforcement for the deck. The other three FRP
deck systems were made entirely out of GFRP and were not filled with concrete. The
26
performance of the deck systems was field tested over a two year period using
controlled truck loads as well as modal tests and results were compared to an adjacent
companion reinforced concrete bridge.
Throughout the monitoring period, the
structural formwork system exhibited composite action with the cast-in-place concrete
deck and performed better than the other GFRP deck systems in terms of impact
factors, thermal effects and system stiffness.
Figure 2.11 Cross section of hybrid FRP panel-concrete bridge deck system with SIP
formwork and tubular stiffeners used in the Salem Avenue Bridge [113]
The performance of the FRP structural formwork system used in the Salem
Avenue Bridge was also experimentally evaluated by Harik et al. using static loading
of single and double span deck slabs with different lengths and depths [114, 115]. The
single and double span panels satisfied serviceability and strength criteria specified by
the Ohio Department of Transportation for consideration in highway bridge
construction. A combined flexure-shear failure mode and debonding of the FRP
tubular sections from the concrete was observed for the hybrid FRP/concrete bridge
deck panels at failure loads between 3.0 and 3.8 times the service loads. Structural
performance was also evaluated under fatigue loading of 10 million total cycles per
deck at extreme temperatures of -30C (-22F) and 50C by Dutta [116] and Lopez-
27
Anido et al. [70]. The hybrid GFRP-concrete system performed favorably in static
and fatigue tests as compared to the all-GFRP decks, and exhibited comparable
performance to the reinforced concrete bridge deck that was tested for comparison.
The exposure of the deck systems to extreme temperatures was observed to cause a
larger effect on performance as compared to the number of applied load cycles.
Further experimental investigations of deck systems using the FRP structural
formwork panels installed in the Salem Avenue Bridge were presented by Dieter et al.
in 2002 [59]. The deck system investigated by Dieter et al., which is illustrated in
Figure 2.12, consists of 203 mm (8 in) thick concrete deck with a pultruded GFRP bidirectional grid for the top reinforcement which was developed by Bank et al. in 1992
[117] as well as pultruded FRP panels with tubular stiffeners and gravel bonded to the
horizontal surfaces. Eight 0.91 m (36 in) wide beams with span lengths ranging from
2.44 m (8 ft) to 2.59 m (8 ft 6 in) and three 2.74 m (9 ft) wide deck panels with span
lengths of 2.44 m (8 ft) and 3.51 m (11 ft 6 in) were tested in either positive or
negative flexure under static and fatigue loading.
Figure 2.12 Hybrid system used in US-151/ Hwy 26 Bridge consisting of pultruded
FRP structural form with aggregate bonded surface and tubular stiffners as well as
FRP grid for top layer of reinforcement (after [67])
28
Detailing modifications were made to later test specimens, based on
observations from testing the first three deck panels, in which bidirectional FRP grids
were placed above the connections between panels in order to reduce longitudinal
cracking above the joints between panels and to improve composite action between
the FRP structural formwork panels and the surrounding concrete [58]. Punching
shear failure was observed in most of the deck test specimens and Dieter
recommended predicting shear capacity using the ACI 318 Equation 11-35 with the
addition of a reduction coefficient, Ag, which takes into account the level of aggregate
bonded to the top surface of the FRP panel [118]. The recommended modified version
of the ACI shear capacity equation is given below:
Vc = Ag 4
f c' bo d
2.1
where
Ag
= reduction coefficient accounts for level of aggregate coverage on top surface
of FRP deck form
= 0.60 for partial coverage
= 1.0 for aggregate bonded to all horizontal surfaces.
'
fc
= specified compressive strength of concrete at 28 days,
bo
= Critical perimeter section of the rectangular loaded area,
d
= distance from the extreme compression fiber to tensile reinforcement,
In 2003, a new bridge on US Highway 151 in Waupun, Wisconsin was

constructed using this bridge deck system. The two-span bridge, with equal spans of
32.7 m (107 ft), was built using the same bridge deck system design examined by
Dieter et al. supported on five prestressed concrete girders spaced at 2.65 m (8 ft 8 in)
on center. Load testing performed on the completed bridge indicated stress levels in
the FRP were within allowable limits [40].
29
Recently, a new stay-in-place structural formwork panel system using
rectangular stiffeners was developed by Cheng et al. [119]. The structural formwork
consists of a bottom plate with foam filled rectangular stiffeners spaced 305 mm (12
in). All surfaces of the panels are treated with a rough sand coating to increase the
level of mechanical interaction between the concrete and the FRP panel. Treatment of
the entire surface with a sand interface is in contrast with many of the other systems
discussed, which provided enhanced bond mechanisms only on the horizontal surfaces
of the panels [65, 69, 71, 73, 97-99, 102, 105]. In addition, the presence of interfacial
shear ribs between each stiffener spaced at 152 mm (6 in) on center, as shown in
Figure 2.13, further improved the mechanical interlock between the cast-in-place
concrete and the FRP panel. Parameters such as the composite layup, the position of
the shear ribs, and the geometry and spacing of the stiffeners were optimized through
experimental and analytical investigations [120].
Figure 2.13 Steel-free hybrid FRP-concrete bridge with FRP tubular girders [119]
The structural formwork panels were developed as part of a steel-free hybrid

FRP-concrete bridge which consists of a polypropylene fiber concrete mix which is
30
cast on top of the FRP structural formwork panels. The panels snap into grooves on
top of hollow FRP tubular girders and a carbon fiber mesh is place above the negative
moment regions, as shown in Figure 2.14. Satisfactory performance was observed for
the 12 concrete slab tests that were performed to characterized component and system
level flexural, shear, and fatigue response [121]. A simplified design procedure and
design guidelines are presented [72].
Figure 2.14 Steel-free hybrid FRP-concrete bridge with FRP tubular girders [119]
Bridges that have been constructed with deck systems containing FRP stay-inplace structural formwork for bridge decks are summarized in Table 2.1.
Table 2.1 Representative bridge systems constructed using FRP structural SIP formwork for bridge decks
31
32
2.3
TYPICAL REPAIR METHODS FOR REINFORCED CONCRETE BRIDGE DECKS
Damaged reinforced concrete bridge decks are repaired using different

methods depending a variety of factors such as the type of damage which occurred, the
availability, cost, structural compatibility and anticipated life of different repair
materials, the environmental conditions near the repair site and availability of
qualified contractors in the area [10, 122]. Chapter 5 describes the repair and retesting
of a test specimen in which damage consisted of cracking of the deck concrete above
the FRP panels and separation at the interface between the concrete and composite
panels.
Therefore, techniques for repairing cracked and damaged concrete are of
primary concern.
For relatively small areas of damage which only extend partially through the
depth of the slab, surface patching using cement-based materials or resin-based
materials can be used [123].
These materials include cement-based mortar or
concrete, non-shrink quick-setting mortar, epoxy mortar, cement-based or resin-based

polymer concrete, and pneumatically applied mortar [122]. The boundaries of the
removal region in the damaged concrete are typically marked by saw cuts and the
unsound concrete is removed to an appropriate depth using either jack hammers or
water jets. After the reinforcing steel is cleaned to the bare metal and reinforcement
with significant deterioration or other forms of damage are replaced, the surface patch
is installed and allowed to cure prior to reopening of the bridge.
Patching has
advantages in terms of cost effectiveness and practicality, however it is generally

considered a temporary repair because patching can cause surrounding steel
33
reinforcement to corrode more quickly due to the differences in chloride ion content
between the patch installation and the adjacent concrete [124].
If a concrete deck is still structurally sound but a significant portion of the deck
surface has deteriorated from heavy in-service traffic, freeze thaw cycles or other
exposures, the deck can be resurfaced with an overlay. A deck overlay provides a
uniform layer of repair material over an extensive area or the entire deck and can also
serve as a wearing agent and moisture barrier. Common material choices for overlays
include low slump high density concrete, latex modified concrete, polymer concrete,
and monolithic resinous toppings [125]. The overlay either rely on full composite
action with the existing bridge deck, which is known as a bonded overlay, or can be
separated from the existing concrete by a flexible interlayer, typically hot-mix asphalt,
to create an unbonded overlay. An unbonded overlay can move independently from
the existing concrete thereby preventing reflective cracking, which is caused by peaks
in strains over cracks from the existing concrete deck. The thickness of the overlay is
25 to 75 mm (1 to 3 in) for a fully bonded overlay and at least 75 mm (3 in) for an
unbonded overly, with the thickness of the overlay based in part on the structures
ability to tolerate the increased dead weight and slab elevation [125].
For cracks in the concrete caused by overloading from an extreme event in
which no further movement is expected, such as those observed in the post-failure test
specimen described in Chapter 5, the cracks may be filled in order to bond together the
concrete surfaces and restore the structural integrity of the system. Cracks larger than
approximately 1mm (0.04 in) can be filled with cement grout, whereas low viscosity
liquid polymers can be used for smaller cracks down to a width of approximately 0.1
34
mm [124]. Cracks filled can be filled by ponding using low viscosity polymer resin
systems under gravity, or through the more common technique of applying positive or
negative pressure to assist the resin to flow into the crack. The resin is injected into a
crack though an injection port which is drilled into the surface of the concrete and the
resin spreads out though the concrete [7].
2.4
2.4.1
NSM FRP SYSTEMS FOR FLEXURAL STRENGTHENING

NSM versus Externally Bonded FRP Reinforcement
The use of FRP composites for flexural strengthening of existing concrete

structures was developed in Europe and Japan in the 1980s as an alternative to steel
plate bonding [126].
Externally bonded FRP systems for flexural strengthening
include but are not limited to wet layup processes with installation on site, resin
impregnation of fabrics and strips, bonding of pre-cured FRP profiles to a structure,
resin infusion of dry fabric after installation of the FRP, and use of prepreg sheets [23,
127]. Application of externally bonded prefabricated strips and externally bonded on
site impregnated fabric laminates for the rehabilitation of bridge deck slabs are shown
in Figure 2.15 (a) and (b) respectively [128].
FRP reinforcement can also be adhesively bonded into precut grooves made in
the cover region of the concrete surface, which is referred to as embedded
reinforcement or near-surface mounted reinforcement [129] rather than bonded to
the surface. Near surface mounted FRP systems only have been actively researched
during the past decade, although the general use of the strategy can be traced to the use
of steel rebar in surface cut grooves in Europe in the 1950s [130].
35
a) Pultruded strips
b) Wet layup fabric laminates
Figure 2.15 Rehabilitation of bridge deck slabs using externally bonded FRP
reinforcement [128]
The use of near-surface-mounted (NSM) FRP reinforcement for rehabilitation

has a number of advantages over the more common externally bonded FRP
reinforcement. These advantages include the potential for reduced site installation
work, since the creation of grooves for the FRP decreases the labor required for
surface preparation, the reduced likelihood of debonding failures from the concrete
surface, and improved protection from mechanical damage provided by recess of the
NSM reinforcement into the concrete surface [129, 131, 132]. Also, prestressing of
FRP reinforcement in order to further utilize the capacity of the FRP material has been
shown to be more effective in NSM applications than externally bonded systems
[133].
The use of near surface mounted FRP rehabilitation techniques provide
particular advantages for flexural strengthening of the negative moment region of

reinforced concrete slabs and decks [129]. In these applications, the top surface of the
deck may be subject to harsh environmental and mechanical conditions, which would
require the FRP reinforcement to be surrounded by a protective cover. An appropriate
36
bond and protection of the FRP reinforcement would be more difficult to achieve
using externally bonded strips whereas the near surface mounted reinforcement is
already embedded and therefore not exposed to these influences [134].
2.4.2
Types of NSM Reinforcement
FRP reinforcement used for near-surface-mounted applications is typically

manufactured using the pultrusion processes and can be manufactured in a wide
variety of shapes including round, oval, square and rectangular bars, as well as strips
with varying aspect ratios [131].
Carbon fiber reinforced polymer composite NSM reinforcement is the primary
type of FRP material used to rehabilitate concrete structures because of the higher
tensile strength and tensile elastic modulus of carbon over glass or aramid [135].
Also, the inertness of the carbon fibers reduces the effect of concrete based alkalinity
on the FRP itself, however the alkalinity still affects the fiber-matrix bond and the
matrix itself. The superior tensile properties of CFRP allow for a smaller crosssectional area NSM CFRP bar to be used over a GFRP or AFRP bar with the same
tensile capacity, which has additional constructability benefits by allowing a shallower
groove to be used, thereby reducing the risk of interfering with the internal steel
reinforcement [131].
While the initial use of NSM was with circular bars, the transition to
rectangular strips was predicated on the desire to achieve better bonding of the
reinforcement [131]. It has been shown that all other factors being equal, near surface
mounted strips have higher average bond strengths than circular bars because the
development of a three-dimensional distribution of bond stresses into the surrounding
37
concrete [136, 137]. Further, in the case of round bars used as near surface mounted
reinforcement, forces due to the radial stresses can induce tensile forces that can force
the bar out of the groove, resulting in bond splitting or failure. Also note that since
strips have significantly larger ratios of perimeter to cross-sectional area than square
or circular rods, bond stresses are lower which allows increased utilization of the
capacity of the NSM strips [137]. While the use of NSM reinforcement for flexural
strengthening is extremely simple, the use of square bars and circular rods requires
larger and deeper grooves than flat strips placed horizontally in order to achieve the
same efficiency.
2.4.3
NSM Failure Modes and Design Assumptions
The distinct failure modes observed for experimentally investigated concrete

structures flexurally strengthened with NSM reinforcement are concrete crushing, FRP
rupture, cover splitting, and bond failure within the concrete adjacent to the adhesive
layer, at the interface between the concrete and the adhesive, within the adhesive
itself, or at the interface between the adhesive and the FRP [126, 129, 134, 138]. A
brief description and explanation of each failure mode as well as relevant design
assumptions are provided below.
2.4.3.1 Concrete crushing

Crushing of the concrete in compression is occurs then the compressive strain
in the concrete reaches its crushing strain. A system which exhibits this failure mode
38
during flexural loading is considered to be over-reinforced.
For design purposes,
crushing of the concrete is assumed to occur at a strain value of cu = 0.003 .
2.4.3.2 FRP rupture

Rupture of the FRP reinforcement occurs then the tensile forces in the FRP
reinforcement exceed the tensile strength of the FRP. Because FRP materials loaded
in tension are observed exhibit linear elastic behavior until tensile failure, FRP rupture
is usually a catastrophic failure mode for the structural element strengthened with
FRP. For design purposes, FRP rupture is assumed to assumed to occur if the tensile
strain in the FRP reaches its design rupture strain, f = fu . In order for FRP rupture
to occur, the system must have the necessary bond strength to delay debonding
failures. While tensile rupture has rarely been observed in NSMR systems which are
not prestressed [134], this failure mode is considerably more likely to occur in systems
which use NSM strips rather than NSM bars or externally bonded FRP (for systems
where all other factors are equal) because the geometry of the NSM strips provides
more efficient bonding to the concrete [138].
2.4.3.3 Bond failures

Debonding occurs when loss of composite action is exhibited across an
interface. Figure 2.16 illustrates the four different bond failure modes that can occur.
Splitting in the concrete adjacent to the concrete-composite interface is a result of the
high tensile stresses that are transferred into the concrete. This failure mode is more
39
common when the adhesive strength is much higher than the concrete strength [139].
A failure at the interface between the concrete and the composite has been reported as
critical for pre-cast grooves due to their even surface [140]. Failure within the
adhesive, known as cohesive failure, can be avoided through the selection of an
appropriate adhesive system. A debond failure between the adhesive and the FRP
reinforcement can be critical for reinforcement with smooth or lightly sand-blasted
surfaces that provide insufficient surface deformation for mechanical interlock
between the reinforcement and the adhesive [137].
Figure 2.16 Bond failure modes for FRP strips (after [134])
2.4.3.4 Cover Splitting

Splitting of the adhesive cover for the NSMR, which is often accompanied by
spalling of the surrounding concrete, is referred to as cover splitting [129]. Cover
splitting is more commonly observed for NSM bars than for strips due to the radial
distribution of bond stresses that load the concrete cover. Conversely, the stress
components for NSM strips act parallels to the concrete surface, so splitting failure is
less likely to occur [134]. NSMR close to the edge of a concrete member can cause
40
edge splitting failure which can be critical, however this failure mode can be avoided
by keeping a minimum distance from the edge members close to the edge [136, 137].
2.4.4
Prior Use
While NSM FRP has been used successfully for flexural strengthening of
concrete beams [46, 137, 141-144] and the flexural strengthening concrete slabs in the
positive moment regions, there is still limited research on NSM FRP applications to
increase the flexural capacity of concrete slabs in the negative moment regions. A
summary of research on flexural strengthening of reinforced concrete slabs with FRP
is presented.
In 1998, Warren reported on the use of NSM FRP bars and strips for increasing
the flexural capacity in negative moment regions of a bridge deck slab [145]. NSM
FRP bars were applied to both reinforced concrete and prestressed concrete slabs and
the NSM FRP bars were used in both the longitudinal and transverse directions on the
slabs. In 2002, NSM FRP bars were used to strengthen the negative moment regions
of a reinforced concrete parking garage in which the steel reinforcement had corroded
[146]. Alkhrdaji et al. reported on the structural response of a decommissioned three
span reinforced concrete bridge deck, in which one slab was strengthened in flexure
using NSM bars [147]. The NSM strengthened slab failed in FRP rupture, whereas
the other slab which was strengthened using externally bonded strips failed by a
combination of FRP rupture and debonding of the sheets. The NSM strengthened slab
was 27% stronger than the unstrengthened control slab, compared to a 17% strength
increase over the control slab for the externally bonded FRP slab.
41
Field studies and laboratory work have shown that NSM FRP strengthening
systems for reinforced concrete structures have significant advantages over externally
bonded FRP strengthening systems. These advantages include enhanced ability to
utilize the FRP prior to debond, less surface preparation required prior to
strengthening, and more efficient and effective material usage. The application of
NSM FRP strengthening to negative moment regions of decks has particular
advantages in terms of ease of construction and better protection for the FRP
reinforcement from the external environment. Research on NSM FRP also clearly
documents the advantages of NSM FRP strips over NSM FRP circular or rectangular
bars. The smaller grooves necessary to mount a strip placed horizontally reduces
concerns about damaging existing reinforcement during groove cutting and accelerates
construction time. The enhanced surface area to volume allows for better utilization
of the FRP reinforcement, which makes FRP rupture at higher loads more likely than
debonding at lower loads. The benefits from NSM FRP strips have been recognized
by the research community, as most recent research has shifted away from NSM FRP
circular and rectangular bars to examination of NSM FRP strips.
2.4.5
Available Specifications for Flexural Strengthening using NSM FRP
The 2nd edition of Concrete Society Technical Report No. 55 discusses a

variety of applications for strengthening with NSM reinforcement in Section 6.4 and
recommends that for aspects other than FRP curtailment, design of flexural
strengthening with NSM reinforcement (NSMR) should be done using the design
methods described for surface mounted reinforcement, with the allowance made to
42
adjust the location of the reinforcement from the surface of the section to within the
section such that the strains in the FRP are lowered appropriately [139]. Approaches
for anchorage design are detailed and design suggestions for reducing the likelihood of
different common modes of failure for NSMR are described.
The previous edition of the ACI 440.2R code, ACI 440.2R-02, made no
specific mention of strengthening using NSMR, however contained extensive
information pertaining to surface mounted reinforcement [148]. The most recent
edition of the ACI 440.2R code, ACI 440.2R-08, has added sections pertaining to near
surface mounted reinforcement.
The new guidelines recommend that flexural
strengthening with NSM FRP systems be designed using the same equations as
externally bonded FRP systems, with different procedures provided for the
determination of the strain level at which debonding may occur (Section 10.1.1),
calculation of the required bar development length (Section 13.3), and detailing of the
bond parameters between the bars and the concrete (Section 13.3) [126].
The 2nd edition of the Canadian Highway Bridge Design Code (CHBDC) for
fiber-reinforced structures includes strengthening with NSMR as part of its discussion
on flexural and axial rehabilitation (Section 16.11.2) and gives resistance factors for
NSMR made from pultruded carbon, glass and aramid (Section 16.5.3) [149]. For
structures with a concrete cover less than 20 mm (0.79 in), NSMR is not permitted
[150]. This code determines NSMR anchorage lengths for flexure using the same
calculation provided for internal FRP bars (Sections 16.11.2.4.4 and 16.8.4.1) and
provides only a general description of failure modes for FRP strengthened systems,
without mention of NSMR specific modes of failure.
USE OF FRP PANELS AS THE STAY-IN-PLACE STRUCTURAL

FORMWORK SYSTEM FOR A BOX GIRDER BRIDGE DECK
3.1
INTRODUCTION
This chapter presents the assessment and validation of a hybrid fiber reinforced
polymer (FRP)-concrete bridge system, where prefabricated composite panels act as
both the formwork and reinforcement for the bridge deck. Experimental results from
the full-scale testing of a box girder bridge test specimen with FRP stay-in-place
structural formwork panels integrated into the bridge deck are discussed.
The
systems structural response as well as the interaction between elements under static
loading is evaluated and a direct performance comparison is drawn between the hybrid
FRP panel-concrete bridge deck system and a conventional steel reinforced concrete
system bridge deck system. The structural aspects addressed in this chapter include
overall structural response, changes in stiffness for the decks, deflection profiles, load
transfer between the panels, strain development, differences in residual deflections,
and progression of damage within the decks.
3.2
OVERALL GEOMETRY AND CONSTRUCTION DETAILS
The overall test specimen geometry consisted of a two-cell box girder bridge
segment, with a center-to-center distance of 1830 mm (6 ft) between each of the stems
and a longitudinal length of 3660 mm (12 ft) as shown in Figure 3.1. The deck slab
was 178 mm (7 in) thick and included overhangs which extended 483 mm (19 in)
beyond the exterior longitudinal edge girders. The bottom slab was 152 mm (6 in) and
43
44
the widths of the outer and center girders were 252 mm (10 in) and 305 mm (12 in)
respectively as shown in Figure 3.2.
Figure 3.1 Overall representation of test specimen
There was also a continuous layer of top steel reinforcement. Traditional

wooden shored formwork and steel reinforcement was used for the remainder of the
specimen. This test setup allowed for a side-by-side performance comparison of the
reinforced concrete deck to the prefabricated FRP panel strengthened system. All
steel reinforcement used within the test specimen was designed in accordance with the
AASHTO-LRFD specifications [91] and the construction practices employed
45
mimicked field techniques. The steel reinforcement in the deck consisted of a top and
bottom layer of 15.9 mm diameter (#5) rebar with the transverse rebar spaced at 203
mm (8 in) on center and variable spacing for the longitudinal rebar in order to
accommodate the location of the girder stems. The top layer of steel reinforcement
was continuous across the test specimen as shown in Figure 3.3, whereas the bottom
layer of steel reinforcement was curtailed completely on the composite side in the
central region due to the presence of the FRP panels which served as the bottom layer
of reinforcement.
The remaining steel reinforcement consisted primarily of 15.9 mm diameter
(#5) rebar for the bottom slab and the stem reinforcement, with two 22.2 mm diameter
(#7) rebar at the bottom of each girder web and 9.5 mm diameter (#3) rebar used for
shrinkage and temperature reinforcement in the longitudinal direction of the girder
webs. Grade 60 steel was used for all the steel reinforcing bars and tensile tests were
performed on samples of the rebar used for construction of the test specimens. The
experimentally determined average tensile strength values of the steel reinforcement
are summarized in Table 3.1. Steel reinforcement detailing for the test specimen is
shown in Figure 3.3 - Figure 3.5.
Figure 3.2 Overall test specimen dimensions- front elevation view
46
47
Figure 3.3 Top layer of steel reinforcement for deck slab- plan view
Figure 3.4 Steel reinforcement detailing in lower portion of box girder - front elevation view
48
Figure 3.5 Steel reinforcement detailing in lower portion of box girderside elevation view
49
50
The specimen was constructed in two separate concrete pours with the initial
pour for the bottom slab and the lower portion of the stems and the second pour for the
deck slab and the upper portion of the stems. Wooden formwork was constructed for
the bottom slab and stems and the reinforcement was detailed such that a clear cover
of 25 mm (1 in) was maintained throughout the specimen. The lower portion of the
specimen was cast with a cement-based concrete that had a nominal design
compressive strength of f c' = 34.5 MPa (5 ksi) at 28 days and an average aggregate
size of 127 mm (0.5 in). The specimen was cured with a moisture barrier on top of the
cast concrete for 7 days prior to formwork removal as shown Figure 3.6.
Table 3.1 Performance characteristics of the Grade 60 reinforcing steel
The strength development plots of all concrete poured for the test specimen, as
obtained from periodic testing of cylinders made during construction of the specimen,
are presented in Figure 3.7 and Figure 3.8 and the results are summarized in Table 3.2.
51
Figure 3.6 Lower portion of the box girder bridge before and after casting of concrete
Table 3.2 Performance characteristics of the concrete
52
45
Concrete Strength (MPa)
40
Truck #1
35
30
25
Truck #2
20
Day of
test
Day of
2nd pour
15
10
5
0
0
14
28
42
56
Days
70
84
98
Figure 3.7 Strength development for concrete from 1st pour

(First pour consists of the bottom slab and the lower portion of the stems)
Compressive strength (MPa)
40
35
30
Day of
test
25
20
15
10
5
0
0
14
21
Days
28
35
42
Figure 3.8 Strength development for concrete from 2nd pour (deck)
After the concrete reached 28 days of cure, the test specimen was placed on top
of the support blocks used as part of the test setup and the deck was constructed in
53
place as shown in Figure 3.9 and Figure 3.10. Details on the construction of the deck
and the installation of the FRP panels are discussed in Sections 3.3 and 3.4.
Figure 3.9 Construction of test specimen formwork for 2nd pour
Figure 3.10 Pouring of deck concrete and finishing of deck surface
54
3.3
FRP PANEL FEATURES
The FRP deck panels used for the test specimens were previously characterized
by Cheng et al. [119, 120] and consist of a bottom plate that serves as the stay-in-place
formwork as well as the bottom layer of reinforcement for the deck, rectangular
stiffeners that span in the direction transverse to traffic loading, and small shear ribs
attached to the bottom plate to increase the shear interlock with the concrete (Figure
3.11) . The bottom plate of the FRP was constructed of hybrid unidirectional carbon
fiber fabric, C, and E-glass continuous strand mat, E, with a lay-up of [C/E/C2/E/C]s.
Rectangular stiffeners with non-crushing lightweight foam cores to maintain the
desired geometry were reinforced primarily with unidirectional carbon fiber fabric.
The foam filler has a density of 0.961 kN/m3 (0.00347 lb/in3), a compression strength
of 0.785 MPa (0.114 ksi), and Youngs modulus in compression of 27.4 MPa (3.98
ksi) [120].
Figure 3.11 FRP deck panel system [119]
55
The stiffeners were bonded to the bottom plate and reinforced with additional
unidirectional carbon fabric which was placed continuously over the panel in the
direction perpendicular to the orientation of the stiffeners. The shear ribs were made
by bonding sections of sand filled epoxy spaced at 150 mm (6 in) to the bottom panel.
A sand coating was applied to the top surface of the panel in order to increase the
surface roughness. Details regarding the layup and geometry of the FRP panels are
shown in Figure 3.12 and Table 3.3.
Figure 3.12 Cross-section details of the stiffeners on the deck panel (after [120])
Table 3.3 Lamina information and nomenclature [120]
The material properties of the FRP panels were previously characterized by

Cheng and these results are summarized below in Table 3.4 and Table 3.5. Table 3.4
provides the experimentally determined properties of the bottom plate (Mark 4) that
56
were obtained from coupon tests of panel samples that which were tested following
the methods prescribed by ASTM Standards [151].
Mark 2: All layers in the top flange and upper web.

Mark 3: All layers comprising the lower web.
Mark 4: All layers in the bottom flange of deck panel.
Mark 6: One layer of the strip.
Table 3.4 Experimental mechanical properties of bottom plate (Mark 4) [119]

Average
Standard
Deviation
T
Tensile Modulus, ET, Longitudinal, EL
MPa [ksi]
Transverse, ETT
60391 [8765]
1883 [273]
6861 [995]
1242 [180]
Longitudinal, ELF
20464 [2970]
2940 [427]
Transverse, ETF
6302 [915]
1375 [200]
Longitudinal, LT
222.5 [32.3]
11.1 [1.61]
Transverse, TT
81.5 [11.8]
25.5 [3.69]
Longitudinal, LF
337.0 [48.9]
59.2 [8.59]
Transverse, TF
89.7 [13.0]
11.4 [1.65]
Longitudinal, L
27.7 [4.03]
1.52 [0.221]
Transverse, T
16.0 [2.32]
1.61 [0.234]
Density, , kN/m3 [lb/in3]
1472 [91.85]
Not available
Nominal Thickness, mm [in]
7.38 [0.291]
0.649 [0.0256]
Direction
Flexural Modulus,
EF, MPa [ksi]
Tensile Strength, T,
MPa [ksi]
Flexural Strength,
F, MPa [ksi]
Interlaminar Shear
Strength, , MPa
[ksi]
Property
The longitudinal direction for the bottom plate was defined as parallel with the
orientation of the stiffeners as opposed to direction of traffic, which was defined as the
57
longitudinal direction for the overall structure.
Table 3.5 shows the mechanical
properties for the different deck panel components, based off of design data.
Table 3.5 Mechanical properties of deck panel components [120]
Mark 2: All layers in the top flange and upper web.

Mark 3: All layers comprising the lower web.
Mark 4: All layers in the bottom flange of deck panel.
Mark 6: One layer of the strip.
58
3.4
PANEL PLACEMENT AND CONNECTION DETAILING
The panels were cut to the desired length of 1830 mm (6 ft) and the cut ends
were sealed with resin to provide a protective moisture barrier for the composite
reinforcement. A central panel with dimensions of 1220 mm (4 ft) wide, 1830 mm (6
ft) long, and 6.3 mm (0.25 in) thick was positioned in the middle of the FRP panel side
of the specimen and was bordered on both sides by narrower panels, each 610 mm (2
ft) wide as shown in Figure 3.12. Multiple panels were used side by side to check the
effectiveness of load transfer between the panels.
In order to accommodate the steel reinforcement protruding vertically from the
girders, the positions of the vertical rebar were measured, and corresponding 25.4 mm
(1 in) holes were drilled in the FRP panels. After the FRP panels were lowered into
place, the steel reinforcement from the girders was bent, as shown in Figure 3.14 (b),
so that the steel reinforcement in the deck could be positioned and tied. Note in Figure
3.14 (a) that wooden shored formwork was installed for the entire deck except for the
region in which the FRP panels are placed. Additional holes were drilled in the end
regions of the panels so that 12.7 mm (0.5 in) diameter threaded rod could be used to
hold the panels in place during the deck concrete pour. The as-built deck prior to the
concrete pour is shown in Figure 3.15.
The times required for each step of construction was monitored and
construction of the FRP panel bridge deck system was determined to take 60% less
time than the construction of the conventional reinforced concrete bridge deck system,
with approximately 25 labor hours versus 63 labor hours required for the construction
of the FRP panel system and RC bridge deck system respectively.
Figure 3.13 Schematic of FRP panel locations

59
60
(a) Composite side deck prior to FRP panel installation
(b) Composite side deck after FRP panel installation

Figure 3.14 FRP panel installation details
61
The installation and removal of the wooden formwork and metal shoring took
approximately 24 labor hours, which was one labor hour less than the time required
for the construction of the entire FRP panel bridge deck system. This is comparable to
other researchers, who reported that the construction of a stay-in-place FRP panel
structural formwork deck system was determined to be 57% faster than the
construction of a comparable steel reinforced deck system with conventional
formwork [67].
Figure 3.15 Plane view looking down on reinforcement system in deck, as built
62
3.5
SPECIMEN SETUP
The test specimen was supported on six 890 kip (200 kN) capacity load cells,
placed directly below the centerline of the three stems and 305 mm (12 in) on-center
inwards from the longitudinal edges of the test specimen as shown in Figure 3.16. The
purpose of these load cells was to evaluate the percentage of loads on the structure
taken by each support and to determine the extent to which the specimen was loaded
evenly throughout the testing.
Figure 3.16 Load cell placement below test specimen
63
Each load cell has a footprint of 254 mm (10 in) by 254 mm (10 in) and
consists of a 51 mm (2 in) thick steel plate placed on top of the load measuring device
with smooth curved surfaces on both faces such that their interaction can be
considered a point contact. The height of the load cells were adjusted in order to allow
the test specimen to sit as level as possible and hydrostone was poured to secure the
position of the load cells so that the load cells read approximately symmetric load
values along both the longitudinal and transverse centerlines when loaded with the self
weight of the test specimen. An illustration of load cells used to support the test
specimen is shown in Figure 3.17.
Figure 3.17 Detail of load cell used to support test specimen
64
Two 980 kN (220 kip) capacity double-rod hydraulic actuators were positioned
1830 mm (6ft) apart from each other and in the center of each span to simulate the
wheels attached to the axle of an AASHTO specified HS20 truck, which has an
unfactored load of 73 kN (16.4 kip) per wheel. In this investigation, all applied load
values are reported as the load applied per actuator, which is half of the total load
applied to the test specimen. A 76 mm (3in) thick steel reinforced elastomeric bearing
pad was placed between the actuators and the specimen and was used to create the 510
mm (10in) by 250 mm (20in) HS20 wheel footprint specified by AASHTO [91].
Figure 3.18 Overall specimen setup
65
3.6
INSTRUMENTATION
The primary measurements taken during the experimental testing of a

structural test specimen are the loads measured by the supports and actuators as well
as the deflections and the strains measured throughout the test specimen. The overall
load distribution within the test specimen was measured using load cells which
supported the structure, deflections were measured with linear potentiometers, strains
were measured using electrical resistance strain gages and rotations were determined
using inclinometers. Table 3.6 summarizes the total quantity of instrumentation used
for this test specimen and the approximate locations in which each instrumentation
type was used.
Table 3.6 Summary of instrumentation used for test specimen #1
66
3.6.1
Linear Potentiometers and Inclinometers
Linear potentiometers were used to measure deflected shape of the structure

under the applied load.
The linear potentiometers were positioned in locations
designed to give an accurate representation of the overall structural response. All of

the linear potentiometers used were attached to stationary support fixtures mounted to
the strong wall and strong floor as shown in Figure 3.18, such that the displacements
measured represented the absolute displacements of the structure.
Due to the positioning of the actuators at the center of the two deck slabs, the
center deflection of each slab needed to be measured from linear potentiometers
mounted below the deck.
For consistency, the deflections throughout the deck slabs
were also measured from linear potentiometers placed on the underside of the deck as
shown in Figure 3.19. Linear potentiometers on the top of the deck were placed along
the centerlines of the three stems to measure the combined deflection of the deck and
stems. Selected linear potentiometers on the top side of the FRP side deck were also
placed in the same position as linear potentiometers located on the underside of the
deck in order to evaluate any separation between the concrete and the FRP panels that
may occur during loading. Two linear potentiometers were placed on the RC side
deck at the same locations as linear potentiometers below the deck in order to establish
the appropriate level of precision which could be afforded when determining whether
any separation had occurred on the FRP side, based on the linear potentiometer pairs.
The theoretical resolution of the linear potentiometers used in testing was
determined based on the analog to digital converter, which has a total of 216 bits,
which is equivalent to 56,536 bits. The theoretical resolution was therefore calculated
67
for a given linear potentiometer as the total stroke of the device divided by 56,536.
The noise level within the system has been experimentally determined to be
approximately 10 times the theoretical resolution and the calculated linear
potentiometer resolutions for the all linear potentiometers used on the test specimen is
presented in Table 3.7. Therefore, changes in deflection with magnitudes that are near
these resolution limit values were considered to be within the noise threshold for the
test setup.
Table 3.7 Resolution attainable by linear potentiometers used on test specimen
68
Figure 3.19 Layout of linear potentiometers located on the underside of the deck
69
Figure 3.20 Layout of linear potentiometers located on the top of the deck
70
A detail of the horizontal linear potentiometers and the inclinometers is shown
in Figure 3.21 and the layout of instrumentation is shown in Figure 3.22. Two linear
potentiometers were mounted horizontally, near the top of the two outer stems,
approximately 76 mm (3 in) longitudinally inwards from the edge of the test
specimen. The three inclinometers used were attached to the outer edge of the test
specimen, with one located at the vertical centerline of each of the three stems and
positioned at mid-plane of the concrete deck slab. The purpose of the horizontal linear
potentiometers and the inclinometers was to measure any changes in slope within the
deck and determine more information regarding the end conditions of the deck slab.
Figure 3.21 Detail of horizontally oriented linear potentiometer and inclinometer
Figure 3.22 Layout of inclinometers and horizontally mounted linear potentiometers

(Inclinometers are attached directly to front surface of test specimen, linear potentiometers are offset inwards 76 mm from edge)
71
72
3.6.2
Strain Gages
Strain gages with 5 mm (0.2 in) gage lengths were attached to the steel
reinforcement on selected rebar in the top and bottom layers of the deck as well as
selected vertical rebar within the stems. Strain gages with 30 mm (1.2 in) gage lengths
were attached to the bottom side of the FRP panel. The precise layouts of the strain
gages are shown in Figure 3.23 through Figure 3.31.
Figure 3.23 Strain gage pattern locations for gages attached to vertical stem rebar
The strain gages attached to the vertical steel reinforcement in the stems are
arranged in one of two patterns, termed A and B, and are located at different
positions within the test specimen cross section as designated in Figure 3.23. Strain
73
gage pattern A contains eight gages, with six attached to the outer stem of the FRP
side and two attached to the center stem as shown in Figure 3.24. Strain gage pattern
B contains three strain gages total, with one strain gage attached to the outer FRP side
stem and two gages attached the center stem as shown in Figure 3.25. There are a
total of 27 strain gages attached to the vertical stem rebar.
Figure 3.24 Strain gage pattern A

(Consists of six strain gages attached to the vertical stem rebar of the FRP side outer
stem and two strain gages attached to the vertical stem rebar of the center stem)
Figure 3.25 Strain gage pattern B

(Consists of one strain gage attached to the vertical stem rebar of the FRP side outer
74
The strain gages attached to the bottom side of the FRP panel side deck were
oriented in the longitudinal and transverse directions, and the positions of these strain
gages are shown in Figure 3.26 and Figure 3.27. The strain gages attached to the
bottom of the FRP deck were spaced 305 mm (1 ft) apart from each other in the
transverse direction and approximately 610 mm (2 ft) apart from each other in the
longitudinal direction.
The longitudinally oriented strain gages were primarily
positioned in two rows along the FRP deck slab, with one longitudinal row at the
midspan of the slab and the other offset from the slab centerline by 305 mm (12 in) in
the direction of the outer stem. These longitudinal lines of strain gages serve to assess
the progression of strains at different distances away from the applied load.
Figure 3.26 Layout of longitudinal strain gages attached to bottom of FRP side deck
75
Figure 3.27 Layout of transverse strain gages attached to bottom of FRP side deck
Strain gages were located directly on either side of the intersection between FRP
panels and on either side of the intersection between the outer FRP panel and the
adjacent concrete in order to evaluate the effectiveness of stress transfer between
adjacent panels and between panels and the surrounding concrete. The transversely
oriented strain gages attached to the bottom of the FRP side deck consisted of one long
row of gages along the test specimen centerline and four smaller rows, located on
either side of the intersection between two panels and on either side of the concrete
FRP panel intersection as shown in Figure 3.27. There were 15 transversely oriented
76
strain gages and 11 longitudinal strain gages for a total of 26 gages attached to the
bottom of the FRP side deck.
Strain gages were also attached to the top and bottom layer of steel
reinforcement in the deck slab and were attached to both longitudinally and
transversely oriented rebar.
Two rows, consisting of three strain gages each were
attached to the top longitudinal rebar that were as close as possible to the centerline of
the two deck slabs. The middle strain gage in each row was placed at the centerline of
the test specimen and the remaining two gages in each row were spaced at 610 mm (2
ft) increments along the bar as shown in Figure 3.28.
Figure 3.28 Layout of strain gages attached to top longitudinal steel reinforcement
77
The longitudinal strain gages attached to the bottom layer of steel
reinforcement consisted of 12 strain gages which were arranged in two rows of three
strain gages each on the RC side and three rows with two strain gages each on the FRP
panel side as shown in Figure 3.29. The rows on the RC side were attached to the
steel rebar which was closest to the centerline of the RC side actuator and the directly
adjacent steel rebar. These gages were spaced using the same pattern as the top
longitudinal gages, with one gage at the test specimen centerline and the other two
spaced at 610 mm (2ft) increments.
Figure 3.29 Layout of strain gages attached to bottom longitudinal steel reinforcement
78
There were 16 total strain gages attached to the top transverse reinforcement,
which were placed in two rows of eight strain gages each. The strain gages were
positioned at 610 mm (2 ft) increments and were attached to the transverse rebar
which ran along the centerline of the test specimen as well as the transverse rebar 813
mm (32 in) away from the specimen centerline as shown in Figure 3.30.
Figure 3.30 Layout of strain gages attached to top transverse steel reinforcement
79
The strain gages attached to the bottom layer of transverse rebar consisted of
six lines of strain gages, which were positioned along the same 610 mm (2ft) spacing
pattern used for the top transverse layer of rebar. Six strain gages were used on the
rebar that was continuous across the specimen, whereas five strain gages were used for
the five rows that contained discontinuous rebar due to the presence of the FRP panels,
which took the place of the bottom longitudinal and transverse steel reinforcement. A
total of 31 gages were attached to the bottom transverse rebar, bringing the total
number of strain gages used to 118.
Figure 3.31 Layout of strain gages attached to bottom transverse steel reinforcement
80
3.7
FORMWORK DEFLECTIONS DURING CASTING OF DECK CONCRETE
Limits on the maximum allowable deflections for stay-in-place formwork

caused by the dead load of the forms, reinforcement and the freshly cast concrete,
serve to prevent excessive sagging during construction.
In order to assess the
effectiveness of the two formwork systems, deflection values due to the construction
of the deck were obtained for the composite deck panel and for the conventional
wooden shored formwork from linear potentiometers (pots) placed directly beneath
the center of each cell.
Concrete was placed to create a 178 mm (7 in) thick deck slab, which applied a
load per unit area of 4.21 kPa (88 psf). During the pouring of the concrete, the
composite deck panel had peak deflections of less than half that of the wooden shored
formwork used for the reinforced concrete side, 2.2 mm (0.087 in) versus 5.5 mm
(0.217 in), respectively as shown in Figure 3.32.
6
Plan View of Deck
Deflection (mm)
5
Conventional
Formwork
4
3
2
FRP Panel
Formwork
Concrete pour
0
0
8
10
Time (hours)
12
14
16
18
Figure 3.32 Comparison of formwork center deflections under construction loading
81
Three days after the completion of the concrete pour, the deflections for the composite
sides FRP panel formwork versus reinforced concrete sides shored formwork were
2.2 mm (0.087 in) versus 5.0 mm (0.197 in) respectively.
While typical formwork deflection values were not available, due to the
numerous variations in wooden shored formwork design, which often depend on
material availability and the preferences of the contractors, AASHTO LRFD Bridge
Design Specifications provide deflection limits for the formwork under construction
loading. For stay-in-place formwork span lengths of less than 3.00 m (10 ft), Section
9.7.4.1 defines the elastic deformation limit under dead load of the forms, plastic
concrete and reinforcement as follows [91]:
max_ form
Lspan
180
= min
12.7 mm (0.5in)
where
max_ form = Maximum allowable deflection of stay-in-place formwork
Lspan
= Span length of the stay-in-place formwork
The conventional shored formwork and the FRP panel formwork both deflected
less than the governing deflection limit of 10.2 mm (0.4 in), which is equivalent to the
span length of 1829 mm (6 ft) divided by 180. These deflection results indicate that
the FRP panels were able to safely act as formwork, exhibiting minimal deflections
under construction loading.
82
3.8
3.8.1
STATIC LOADING OF TEST SPECIMEN

Loading Protocol
The loading values used in experimental testing were based on the HS20 truck
wheel loading described in the AASHTO Bridge Specifications [91], where the
vehicle load was the only load effect considered. Two 980 kN (220 kip) capacity
double-rod hydraulic actuators were positioned 1830 mm (6ft) apart from each other
and in the center of each span to simulate the wheels attached to the axle of the HS20
truck, which has an unfactored load of 73 kN (16.4 kip) per wheel.
In this
investigation, all load values are reported as the load applied per actuator as opposed
to the total load applied to the test specimen. Throughout the testing, two equally
loaded actuators were used, therefore the total applied load would simply be twice the
load applied per actuator. In order to ensure that the actuators remained in contact
with the specimen during testing, a minimum threshold load value of 4.5 kN (1 kip)
was applied to the test specimen for the unloaded state during static loading.
For static loading, the service load was defined by the AASHTO SERVICE I
level (clause 3.4.1 and Table 3.4.1-1 of the AASHTO specifications) in which the
HS20 wheel load of 73 kN (16.4 kip) with an additional AASHTO defined 33%
dynamic allowance for the static loading case, or 98 kN (22 kips) was applied [91].
Load levels were also compared with the AASHTO STRENGTH I level, which was
computed as the SERVICE I level with an additional 75% live load factor and with the
CALTRANS specified Permit truck load of 180 kN (41 kip), which is calculated from
a 106 kN (24 kip) wheel load with a 30% dynamic allowance and additional 30% live
load factor [152].
83
The test specimen was loaded statically to failure with single load/unload
cycles at increasing load levels up to the ultimate load of 890 kN (200 kips). The first
load level applied was the AASHTO service load for static testing of 98 kN (22 kips)
as described above. Next, the specimen was loaded to 222.4 kN (50 kips), cycled
through the minimum threshold level of 4.5 kN (1 kip) and then loaded in increments
of 89 kN (20 kip). Initial signs of punching shear were observed at 823 kN (185 kips),
which dictated that the test specimen be loaded more slowly until the ultimate capacity
of 890 kN (200 kips) was reached. A summary of the static loading protocol is shown
in Table 3.8 .
Table 3.8 Static loading protocol
84
3.8.2
Overall Behavior
At each load level in which a single cycle was performed, the target load was
held briefly to allow inspection of the test specimen in the loaded state prior to
unloading the specimen. This allowed for observation of any cracking or damage
which had occurred to the test specimen during a given load level. Indications of
partial separation between sections of the FRP panels and the concrete were first
observed in the strain and deflection data at the 489 kN (110 kip) load level, which
was equivalent to 5 times the AASHTO service load and is discussed in Section 3.8.8,
however no cracking or other visible damage was detected during an inspection of the
deck surface at this load level. At the 578 kN (130 kip) load level, it was noted that
the elastomeric bearing pad below the composite actuator had shifted horizontally by a
approximately 5 millimeters (0.2 in) from its original location.
Discontinuous cracking was first observed on the top surface of the deck at the
667 kN (150 kip) load level with the most significant cracks located in the longitudinal
direction above the central stem, followed by longitudinal cracking above the RC side
stem and only hairline longitudinal cracks above the composite side stem. By the 756
kN (170 kip) load level, discontinuous longitudinal cracks that ran along the length of
the specimen above each stem were observed. Some cracking was also noted near the
center of the bottom side of the RC side deck in the longitudinal and transverse
directions at this load level.
The FRP panel reinforced slab failed in punching shear at a load of 890 kN
(200 kip), which is equivalent to 9.1 times AASHTO service load, with initial signs of
punching shear observed at a load of 825 kN (186 kip), which is equivalent to 8.4
85
times AASHTO service load. The midspan deflections at the maximum applied load
were 7.09 mm (0.279 in) for the FRP panel reinforced slab and 7.81 mm (0.307 in) for
the reinforced concrete slab. The overall load versus midspan displacement response
curves for the two slabs are shown in Figure 3.34. In order to better compare the
structural response of the two reinforcing schemes, the unloading portion of each
curve was removed and the envelope plot of load versus center deflections is shown in
Figure 3.33.
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
RC side
Load (KN)
Composite side
3
4
Deflection (mm)
Figure 3.33 Envelope plot of load versus center deflections
At 98 kN (22 kip), equivalent to the AASHTO HS20 wheel load of 73 kN

(16.4 kip) with an additional 33% impact factor, the vertical midspan deflection for
both slabs was 0.29 mm (0.011 in). The midspan deflections of the two slabs were
within 0.1 mm (0.004 in) of each other up to a load level of 338 kN, equivalent to 3.4
times the AASHTO service load.
1000
Composite side
RC side
9x AASHTO 900
service load
800
Load (kN)
700
600
500
400
300
200
1x AASHTO 100
service load
0
0
3
4
Deflection (mm)
Figure 3.34 Load versus center deflections

86
87
For the remainder of the loading, the reinforced concrete side deflected more at the
midspan than the composite side, with a maximum difference between the two slabs of
1.5 mm (0.59 in), equivalent to a 31.0% greater RC side deflection, which occurred at
a load of 762 kN (171.3 kip). The greatest percentage difference between the two
sides, in which the reinforced concrete side deflected 46.6% greater, occurred at a load
of 497 kN and corresponded to a 2.32 mm (0.091 in) composite side deflection and a
3.4 mm (0.134 in) reinforced concrete side deflection.
3.8.3
Stiffness Change Throughout the Loading Range
In order to compare the progression in deterioration of performance of the

slabs as observed through degradation of the slope of load-deflection response of the
slabs over the load cycles, a parameter known as the effective stiffness, was defined
as follows:
Effective Stiffness =
Pf Pi
f i
3.1
where
Pf
= Target actuator load of a load cycle based on loading protocol
Pi
f
i
= Minimum threshold actuator load of a load cycle

= Deflection at load Pf , at the location under consideration
= Deflection at load Pi , at the location under consideration
Note that Equation 3.1 uses the minimum threshold actuator load of
approximately 4.5 kN (1 kip) and the corresponding deflection at the start of a load
88
cycle, Pi and i , in the effective stiffness calculation. Conversely, a standard secant
stiffness calculation uses either the origin or the minimum threshold load and
deflection value at the start of experimental testing for the lower value.
This
modification from a standard secant stiffness calculation was chosen in order to

evaluate the degradation of the slope of the load-deflection curve independent of the
effects of accumulated residual deflections on the stiffness results.
The results
obtained from the chosen calculation method shown in Equation 3.1 termed the
adaptive secant stiffness method are compared with results obtained using the
standard secant method and a discussion on this topic is found later in this section.
Also note that the effective stiffness obtained from the load-deflection
response is different from the actual bending stiffness of the slab because the
calculation does not take into account the moment curvature response. The effective
stiffness data calculated can be used for comparative evaluation of the relative
progression of change in structural response over the load cycles.
In order to examine changes in response, the effective stiffness ratio rather than
absolute values of effective stiffness are presented in this section. The effective
stiffness ratio was calculated by comparing the effective stiffness at a load level to the
effective stiffness at the baseline load as shown in Equation 3.2:
Effective Stiffness Ratio =
Effective stiffness at a load level

Effective stiffness at baseline (98 kN)
3.2
The load of 98 kN (22 kip), which represented a HS20 truck load of 73 kN

(16.4 kip) with an additional AASHTO defined 33% dynamic allowance and was also
89
the initial load level in the static testing, was chosen as the baseline load. Since the
response of the test specimen was in the linear elastic range at this load level, the
effective stiffness ratio at this load level is taken to be 1.0.
The effective stiffness ratios shown in Table 3.9 and Table 3.10 highlight a
significant difference in the degradation trends of the load deflection slopes for the
two reinforcing systems. The degradation trends for the two slabs are most easily
observed in Figure 3.35, which plots the effective stiffness values of both slabs for
each load level.
1.0
Effective Stiffness Ratio
Composite side
0.8
RC side
0.6
0.4
0.2
0.0
98 222 311 400 489 578 667 756 845 890
Load Level (kN)
Figure 3.35 Effective stiffness ratios of the two slabs determined from linear
potentiometer data
90
Table 3.9 Effective stiffness ratios of FRP side slab from linear potentiometer data
Table 3.10 Effective stiffness ratios of RC side slab from linear potentiometer data
91
The reinforced concrete side slab displayed a 17% degradation of the effective
stiffness between the first and second load levels and then continued to decrease in the
slope of the load-deflection response for the next four load levels. These first six load
levels, which loaded the specimen up to 578 kN (130 kip) (equivalent to 5.9 times the
AASHTO service load) accounted for over 85% of the total effective stiffness
degradation that occurred in the reinforced concrete side slab during testing and only
minor changes in effective stiffness ratio occurred during the remaining four load
levels.
The FRP panel reinforced side slab exhibited a significantly higher effective
stiffness ratio in comparison to the reinforced concrete slab throughout the loading
range and maintained a relatively constant effective stiffness ratio of 0.9 or greater for
the first five loading levels. At the 578 kN (130 kip) load level, which is 5.9 times the
AASHTO service load, the effective stiffness ratio of the composite side decreased by
8% to 0.83 indicating that additional damage had occurred at this load level. The
damage that occurred at this load level however, did not significantly affect the loaddeflection slope performance of the system, which decreased an average of 5 % per
level during the next two load levels. At the 845 kN (190 kip) load level, equivalent to
8.6 times the AASHTO service load, a 13% decrease in the effective stiffness ratio to
a value of 0.65 was observed. Note that initial signs of punching shear were observed
on the FRP panel side slab at this load level.
For comparison, the effective stiffness and stiffness ratio values were
recalculated with the secant method using the following equation:
92
Effective Stiffness =
Pf Po
f o
3.3
where
Pf
Po
f
= Nominal actuator load at the start of experimental testing

= Deflection at target actuator load, Pf , at the location under consideration
= Deflection at nominal actuator load, Po , at the location under consideration
Note that the nominal minimum threshold actuator load of 4.5 kN (1 kip) at the
start of experimental testing was used throughout the effective stiffness calculation as
opposed to the nominal load and deflection at the start of each load cycle as described
previously in Equation 3.1.
By using the same lower value for the stiffness
calculation, the effect of the residual deflections that occur between load cycles is
taken into account, therefore these calculated effective stiffness values are lower than
the previously calculated values which ignored residual deflections. The difference
between the two calculation methods for the effective stiffness of the bridge decks is
illustrated in Figure 3.36.
A comparison between the effective stiffness ratios determined using the two
different load-deflection slope calculation methods is given in Table 3.11, in which the
effective stiffness results as well as the percent difference between the two methods
are tabulated. The first two load level of the effective stiffness ratios are the same for
both methods because the same nominal load and deflection value is used in both
calculation methods whereas the adaptive secant stiffness method utilizes different
nominal load and deflection values for all higher load levels and yields comparatively
larger effective stiffness ratios for the remaining load levels.
93
(a) Secant stiffness method (Eqn 3.3) (b) Adaptive secant stiffness method (Eqn 3.1)
Figure 3.36 Illustration showing difference in stiffness calculation methods
Table 3.11 Comparison of effective stiffness ratios found using two different methods
of calculating effective stiffness
94
In field applications of a bridge deck system, desirable load-deflection slope
characteristics of the deck both with and without residual deflections included are very
important. Under a given traffic load, the total deflection of the bridge including the
accumulated residual deflections determines whether a system can meet the
serviceability requirements regarding maximum live load deflections specified by
AASHTO. However, the quality of ride for motorists and pedestrians crossing the
bridge is established from the effective stiffness calculations which determine the
slope of the load deflection curve without including the residual deflections. Figure
3.37 graphically summarizes the data from Table 3.11 to illustrate the difference
between the two methods for calculating the effective stiffness.
1.0
composite sidemodified secant
0.8
composite sidesecant
0.6
0.4
RC sidemodified secant
0.2
RC sidesecant
0.0
222
400
578
Load Level (kN)
756
Figure 3.37 Effective stiffness ratios for two methods of calculating stiffness
95
Based on the effective stiffness ratios calculated using both the secant method
described on page 92 and the adaptive secant stiffness method described on page 87,
the composite side slab exhibits significantly more desirable load-deflection slope
characteristics than the reinforced concrete side slab. The effective stiffness ratios
determined using the adaptive secant stiffness method (Equation 3.1) show that for
load levels up to the 489 kN (110 kip), which is equivalent to approximately 5 times
the AASHTO service load, the slope of the load-deflection curve for the composite
side only degraded by 10% overall whereas the reinforced concrete side exhibited a
38% overall decrease in the effective stiffness ratio (Figure 3.35). Additionally, the
composite side maintained larger effective stiffness ratio values as compared to the
reinforced concrete side for all load levels and degraded at a significantly slower rate.
The composite side also retained larger effective stiffness ratios throughout the
loading range for the secant stiffness calculations which included the residual
deflections of the bridge decks.
3.8.4
Deflection Profiles
In this section, deflection profiles throughout the test specimen for key load
levels up to ultimate capacity are presented. The primary purpose for obtaining the
displacement profiles was to determine the extent to which the two reinforcement
systems were able to distribute the wheel load over a larger area in both the
longitudinal and the transverse directions, thus helping to avoid localized punching
shear failure. Unless otherwise noted, the deflection profile figures in this section all
maintain the same deflection scale, which accommodates the maximum values
96
observed in the specimen at ultimate capacity, in order provide context for each
deflection value in relation to the overall deformation of the structure.
The deflection profiles along the length of the deck above the central and outer
stems are presented in Figure 3.38, Figure 3.39 and Figure 3.40. These figures
indicate that the deck above each of the three stems deflected evenly along the length
of the stems. The uniform deflections observed in deck above the stems indicate that
the vertical deformations of the stems do not significantly affect the overall structural
response of the test specimen.
Deflection (mm)
7
6
Max slab
deflection
(890 kN)
+
"0"
5
4
3
2
1
0
-1830
890 kN
578 kN
400 kN
98 kN
-1220
-610
0
610
1220
Position from center of actuator (mm)
1830
Figure 3.38 Deflection profile along length of deck above central stem
The maximum deflection in the deck above the stems at ultimate capacity was
1.3 mm (0.05 in), which indicates that the actual corresponding maximum deflections
at the midpoints of the slabs are approximately 1 mm (0.04 mm) smaller than the
values given from the linear potentiometers due to the deflection contribution from
97
stems. Note that the maximum deflections above the stems at ultimate capacity were
approximately 6.2 times smaller than the corresponding midpoint deflection of the
reinforced concrete side slab.
8
7
Deflection (mm)
Max slab
deflection
(890 kN)
+
"0"
5
4
3
2
1
0
-1830
890 kN
578 kN
400 kN
98 kN
-1220
-610
0
610
1220
1830
Figure 3.39 Deflection profile along length of deck above composite side stem
8
7
Deflection (mm)
Max slab
deflection
(890 kN)
+
"0"
5
4
3
2
1
0
-1830
890 kN
578 kN
400 kN
98 kN
-1220
-610
0
610
1220
1830
Figure 3.40 Deflection profile along length of deck above RC side stem
98
The deflection profiles along the centerline of the two slabs are shown in
Figure 3.41 and Figure 3.42. Both slabs maintained comparably shaped deflection
profiles with larger deflections observed throughout the reinforced concrete side for
loadings up to 756 kN (170 kip), equivalent to approximately 7.7 times the AASHTO
service load.
For the remaining loading up to ultimate capacity, the reinforced
concrete side continued to deflect to a greater extent whereas the composite side
exhibited a more localized deformation profile at these higher load levels.
As
expected, the highest deflection values for both slabs were observed from the linear
potentiometers below the two hydraulic actuators, with progressively smaller
deflections at distances away from these locations. Similarly, because the centerline
deflection profiles shown in Figure 3.41 and Figure 3.42 were obtained from locations
directly below the actuators, these figures contain higher deflection values as
compared to deflection profiles obtained from locations offset from the centerlines,
shown in Figure 3.50 - Figure 3.53.
In order to more thoroughly compare the level of localized deformations in the
slabs, the superimposed deflection profiles along the centerline of the decks are shown
at different load levels. The deflection profiles shown in Figure 3.43 at the load level
of 756 kN (170 kip), equivalent to 7.7 times the AASHTO service load, are
comparably shaped for the two sides, with larger deflections along the entire length of
the reinforced concrete side.
99
8
Max slab
deflection
(890 kN)
Deflection (mm)
+
"0"
890 kN
823kN
756 kN
578 kN
400 kN
222 kN
98 kN
3
2
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 3.41 Deflection profile along centerline of composite side deck
8
Max slab
deflection
(890 kN)
Deflection (mm)
+
"0"
890 kN
823kN
756 kN
3
578 kN
2
1
0
-1830
400 kN
222 kN
98 kN
-1220
-610
610
1220
1830
Figure 3.42 Deflection profile along centerline of RC side deck
100
At the 845 kN (190 kip) load level, which is equivalent to 8.6 times the
AASHTO service load and was the load level at which initial signs of punching shear
damage were observed, the composite side exhibited more localized deflections in the
close proximity to the actuator, however still maintained smaller deflections overall
than observed in the reinforced concrete side as shown in Figure 3.44. At ultimate
capacity, as shown in Figure 3.45, the localized deflections near the point of load
application were significantly more pronounced for the composite side, however, the
deflections were still less than the corresponding reinforced concrete values and the
load level was equivalent to over nine times the AASHTO service load.
Max slab
deflection
(890 kN)
Deflection (mm)
7
6
5
4
RC side
3
2
+
"0"
Composite
side
1
0
-1830
-1220 -610
0
610
1220
1830
Figure 3.43 Comparison of deflection profiles at a load level of 756 kN
101
8
Max slab
deflection
(890 kN)
Deflection (mm)
7
6
+
"0"
5
4
RC side
Composite
side
2
1
0
-1830
-1220
-610
0
610
1220
1830
8
6
Max slab
deflection
(890 kN)
RC side
Deflection (mm)
+
"0"
4
3
Composite
side
2
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 3.45 Comparison of deflection profiles at ultimate capacity (890 kN)
102
An alternative comparison of the deflection values along the centerline of the
two spans was performed by calculating the percentage change in deflection between a
composite side linear potentiometer and a reinforced concrete side linear
potentiometer located at the same distance along the centerline of the deck slabs. The
percentage change in deflections values were calculated with respect to the composite
side and were computed at different load levels. The results for locations 305 mm (2
ft) on either side of the center deflections (Figure 3.47 and Figure 3.48) and 610 mm
(4 ft) away from the center deflections (Figure 3.47) all show comparable trends for
the different load levels.
The percent change in deflections between the two sides at these locations
increases steadily with successive load levels up to a load level of 845 kN (190 kip),
which is equivalent to 8.6 times the AASHTO service load, followed by a decrease in
values for the ultimate capacity load level of 890 kN (200 kip). This trend illustrates
that as the load to which the test specimen is subjected increases, the deflection values
on the reinforced concrete side at these locations become increasingly greater than the
composite side up until ultimate capacity is reached where the difference between the
deflections both sides becomes less pronounced. In contrast, comparison of central
deflections shown in Figure 3.49 illustrates a steady increase in values up to a load
level of 489 kN (110 kip) followed by relatively constant values for the subsequent
four load levels and then a decrease in values for the 845 kN (190 kip) load level and
for the ultimate capacity load level.
The difference in the response at the center points as compared to the other
locations along the centerlines of the two slabs can be attributed to the more localized
103
loading of the FRP panel in the longitudinal direction.
Note that although the
composite side displays a slightly more localized deflection response in the

longitudinal direction, the reinforced concrete side deflects significantly more at all
locations and at all load levels examined.
For instance, at the 489 kN load level,
which is equivalent to approximately five times the AASHTO service load, the
reinforced concrete side deflected between 23% and 42% greater than the composite
side deflections for the linear potentiometers along the longitudinal profile whereas at
the 845 kN load level, the reinforced concrete side exhibited between 24% and 73%
greater deflections.
% Change in deflection values
80%
70%
60%
50%
40%
30%
Reference Comparison
point
point
20%
10%
0%
98 222 311 400 489 578 667 756 845 890
Load Level (kN)
Figure 3.46 Percent change between composite side and RC side deflection values at
different load levels- linear potentiometers located at 2/3 span along the centerline of
each deck
104
80%
70%
60%
50%
40%
30%
point
point
20%
10%
0%
98 222 311 400 489 578 667 756 845 890
Load Level (kN)
different load levels- linear potentiometers located at 1/3 span along the centerline of
each deck
80%
70%
60%
50%
40%
30%
20%
Comparison
point
Reference
point
10%
0%
98 222 311 400 489 578 667 756 845 890
Load Level (kN)
Figure 3.48 Percent change between composite side and reinforced concrete side
deflection values at different load levels- linear potentiometers located at 1/3 span
along the midspan of each deck
105
80%
70%
60%
50%
40%
30%
point
point
20%
10%
0%
98 222 311 400 489 578 667 756 845 890
Load Level (kN)
different load levels- linear potentiometers located at the center of each deck
The deflection profiles along the length of the deck, which are offset from the
centerlines of the slabs by 305 mm (12 in), are shown in Figure 3.50 - Figure 3.53 and
give useful information regarding the effectiveness of the reinforcement systems to
distribute the actuator load to the surrounding regions. For load levels up to 578 kN
(130 kip), the deflection profiles in the longitudinal direction near the outer stems
(Figure 3.50 and Figure 3.51) as well as the profiles near the central stem (Figure 3.52
and Figure 3.53) illustrate comparable deflections in both slabs, with slightly larger
values observed on the reinforced concrete side. For the remaining load levels until
ultimate capacity, the composite side clearly exhibits the same localized deformation
in close proximity to the actuator that was noted for the centerline deflection profile in
Figure 3.41.
106
Deflection (mm)
7
6
Max slab
deflection
(890 kN)
890 kN
+
"0"
5
4
823kN
756 kN
578 kN
400 kN
222 kN
98 kN
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 3.50 Deflection profile along row 2 of linear potentiometers

(offset from composite side centerline by 305 mm)
8
7
Deflection (mm)
6
5
4
3
2
1
0
-1830
Max slab
deflection
(890 kN)
890 kN
823kN
+
"0"
756 kN
578 kN
400 kN
222 kN
98 kN
-1220
-610
0
610
1220
1830

(offset from RC side centerline by 305 mm- see Figure 3.19)
107
Deflection (mm)
7
6
5
4
3
2
1
0
-1830
Max slab
deflection
(890 kN)
890 kN
823 kN
+
"0"
756 kN
578 kN
400 kN
222 kN
98 kN
-1220
-610
0
610
1220
1830

(offset from composite side centerline by 305 mm- see Figure 3.19)
8
7
Deflection (mm)
6
5
4
3
2
1
0
-1830
Max slab
deflection
(890 kN)
890 kN
823kN
+
"0"
756 kN
578 kN
400 kN
222 kN
98 kN
-1220
-610
0
610
1220
1830

(offset from RC side centerline by 305 mm- see Figure 3.19)
108
The deflection profile along the transverse centerline of the test specimen is
presented in Figure 3.54.
The highest deflection values were observed near the
actuator locations and minimal deflection values were observed above each of the
three stems. Both sides maintained comparable deflection profiles up to the 400 kN
(89.9 kip) load level, equivalent to 4.1 times the AASHTO service load. At the higher
load levels, the reinforced concrete side exhibited significantly more localized
deflections in close proximity to the point of load application as shown in Figure 3.54,
which is opposite the trend observed for the longitudinal direction deflection profile
where the composite side exhibited more localized deflections at higher load levels.
The significantly improved distribution of the localized point load for the composite
side slab transverse deflection profile as compared to the longitudinal deflection
profile can be attributed to the presence of the stiffeners described in Section 3.3,
which are oriented in the transverse direction.
In addition to the centerline profile shown in Figure 3.54, the displacement
profiles of three transverse rows of linear potentiometers are presented in Figure 3.55,
Figure 3.56 and Figure 3.57, which are offset from the centerline by 1525 mm (60 in),
1245 mm (49 in) and 635 mm (25 in) respectively. The deflections observed on the
reinforced concrete side at load levels above 400 kN (89.9 kN) are consistently higher
than the composite side deflections.
Max slab
deflection
(890 kN)
"0"
Deflection (mm)
890 kN
5
823kN
4
756 kN
578 kN
400 kN
1
222 kN
98 kN
0
0
610
1220
1830
2440
3050
Position from composite side edge of deck (mm)
3660
4880
109
Figure 3.54 Deflection profile along centerline of specimen
4270
110
Max slab
deflection
(890 kN)
Deflection (mm)
"0"
890 kN
823 kN
756 kN
578 kN
400 kN
222 kN
98 kN
4
3
2
1
0
0
610 1220 1830 2440 3050 3660 4270

Figure 3.55 Deflection profile along row A of linear potentiometers

(offset from test specimen centerline by 1525 mm- see Figure 3.19 and Figure 3.20)
Max slab
deflection
(890 kN)
Deflection (mm)
"0"
5
4
890 kN
823 kN
578 kN
756 kN
400 kN
222 kN
98 kN
3
2
1
0
0
610 1220 1830 2440 3050 3660 4270

Figure 3.56 Deflection profile along row B of linear potentiometers

111
8
Max slab
deflection
(890 kN)
Deflection (mm)
7
6
"0"
5
890 kN
823 kN
578 kN
4
3
756 kN
400 kN
222 kN
98 kN
2
1
0
0
610 1220 1830 2440 3050 3660 4270

Figure 3.57 Deflection profile along row C of linear potentiometers

The envelope plots of load versus deflection for linear potentiometers adjacent
to the central linear potentiometer locations are shown in Figure 3.58 and Figure 3.59.
Figure 3.58 compares the load-deflection plots of the linear potentiometers that are
located a distance of 610 mm (2 ft) away from the center of the test specimen along
the centerlines of the two slabs. The very similar curves for linear potentiometers the
two linear potentiometers on the composite side (C3B and D3B) indicate a very
symmetric deflection profile for the composite side deck in the longitudinal direction
whereas the reinforced concrete side exhibits slightly unsymmetrical deflections (C7B
and D7B). For the composite side, the load-deflection curves remain approximately
linear up to the 756 kN (170 kip) load level and displays a more relaxed slope for the
remainder of the loading. Conversely, the reinforced concrete side only maintains an
112
approximately linear slope up to the 311 kN (70 kip) load level and then exhibits a
nonlinear response for the remainder of the loading.
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
C3B D3B C7B D7B

C7B
D3B
D7B
Load (KN)
C3B
3
4
5
Deflection (mm)
Figure 3.58 Envelope plot of load versus deflections for linear potentiometers along
the centerlines of the two slabs
(located 610 mm from the test specimen centerline- see Figure 3.19)
The load versus deflection envelope plots for linear potentiometers which lie
along the centerline of the test specimen and are positioned 610 mm (2 ft) on either
side of the middle of both deck slabs are shown in Figure 3.59. Note that the linear
potentiometers at three of the four locations (M4B, M6B and M8B) display equivalent
envelope plots while the fourth linear potentiometer (M2B), located on composite side
slab adjacent to the exterior stem, exhibits a higher slope and smaller deflections
throughout the loading range. The deflection values for linear potentiometer M2B
during experimental testing were an average of 20% smaller than the deflections at the
other three locations with a standard deviation of 4%.
113
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
M2B
M4B
M6B
Load (KN)
M8B
M2B M4B M6B M8B

0
3
4
5
Deflection (mm)
Figure 3.59 Envelope plot of load versus deflections for linear potentiometers along
the centerline of the test specimen
(located 610 mm on either side of the deck centerlines- see Figure 3.19)
The maximum deflection values measured in the deck slab were located at the
midspan of the two slabs, directly below the two actuators. At the 890 kN (200 kip)
load level, equivalent to over nine times AASHTO service load, the composite side
displayed a 7.09 mm (0.279 in) center deflection, while the reinforced concrete side
displayed a 7.81 mm (0.307 in) center deflection.
A detailed discussion of the
progression of the center deflections, as they pertain to the effective stiffness of the
slabs, is provided in Section 3.8.3, beginning on page 87 for the FRP panel reinforced
slab and for the reinforced concrete slab.
One important difference between the deflection profiles in the longitudinal
direction versus the transverse direction is the orientation and design of the stiffeners
and the layup of the FRP panels. Due to the transverse orientation of the stiffeners,
the transverse direction deflection profile is significantly improved due to the
114
additional stiffening added by the stiffener elements while the longitudinal direction
deflection response of the composite side deck slab is not impacted considerably by
the presence of the stiffeners. In comparison to the reinforced concrete side, the
transverse direction deflection profile for the composite side is significantly more
dispersed as observed in Figure 3.54 and the concrete side deflects much more in the
localized region surrounding the actuator, especially at high load levels.
In the
longitudinal direction, in which the stiffeners do not play a large role in the structural
response, the two systems exhibit comparable deflection profiles, with the reinforced
concrete side deflecting more throughout and the composite side exhibiting a slightly
more localized response at high loading levels.
3.8.5
Strain Development
The strain development within the test specimen was observed though the
examination of strain gages applied to the steel reinforcement as well as to the FRP
panels within the test specimen. The precise location of all strain gages is given in
Section 3.6.2. Note that strain values reported as negative indicate compressive strains
whereas positive values indicate tensile strains.
As discussed in Section 3.1, the FRP structural formwork panels take the place
of the bottom steel reinforcement. Since there is no bottom steel reinforcement for the
central section of the composite side deck, the strain measurements for the bottom of
the slab are obtained from strain gages attached to the bottom of the FRP panels.
Conversely, strain gages attached to the lower level of steel reinforcement in the deck
slab are used for the remainder of the test specimen. The difference between the
115
position within the depth of the deck slab for the strain gages attached to the
composite side and reinforced concrete side deck slabs is approximately 41.3 mm
(1.63 in) for the bottom transverse strain gages and 57.2 mm (2.25 in) for the bottom
longitudinal strain gages as illustrated in Figure 3.60. The total thickness of the deck
slab is only 178 mm (7.0 in), therefore a difference in position of 57.2 mm (2.25 in)
accounts for over 30% of the total thickness of the slab. As a result, these strain
values cannot be directly compared without further data manipulation to account for
the difference in positions.
Transverse rebar
Strain gage
Longitudinal rebar
41.3 57.2
(a) RC side bottom transverse strain gage on top of bottom layer of transverse rebar
Longitudinal rebar
Transverse rebar
FRP panel
Strain gage
(b) Composite side bottom transverse strain gage on bottom of FRP panel
Figure 3.60 Deck slab cross-section illustrating location of strain gages within depth
(measurements, in millimeters, show the difference in depth between strain gages
attached to FRP panel and strain gages attached to bottom steel reinforcement)
116
In order to more easily compare the strains in the FRP panel to the strains in
the bottom steel reinforcement, a linear strain distribution as shown in Figure 3.61 was
assumed. The modification in strain values is done by adjusting the strains measured
in the bottom layer of steel reinforcement on the reinforced concrete side along an
assumed linear strain distribution to the equivalent depth within the slab where the
FRP panel strain gages were attached on the FRP panel side deck slab. The predicted
strains are likely to display larger variations from actual strains for regions of
discontinuity within the specimen, however the linear strain distribution and
implemented calculations described below have yielded good correlation with
experimental results for segments where the FRP panel and bottom steel reinforcement
overlap, thereby allowing direct validation of predicted strain values.
s
s*
Composite side cross-section
Strain distribution
RC side cross-section
Figure 3.61 Illustration of strain distribution and adjustment of RC side strains
The experimentally determined bottom layer rebar strains, s , were adjusted

along the linear strain distribution profile to the same height within the deck slab as
the FRP panel strain gages, with the adjusted steel strains labeled as s* .
The
curvature, , was determined using a straight line theory in which a linear stress
117
distribution was assumed and transverse effects as well as the top reinforcement were
ignored [153]. The strain line theory calculation method uses the reinforcement ratio,
, and the modular ratio, n , which are defined in Equations 3.4 and 3.5 respectively.
where
As
b
d
Es
Ec
As
bd
3.4
n=
Es
Ec
3.5
= Area of tensile steel reinforcement

= Section width
= Distance from the extreme compression fiber to tensile reinforcement
= Youngs modulus of steel
= Youngs modulus of concrete
The values for and n are used to determine the effective length factor for
compression members, k , as calculated in Equation 3.6. This term locates the neutral
axis at a distance kd down from the extreme compression fiber of the section as
shown in Figure 3.62.
k = 2 n + ( n ) n
2
where
= Reinforcement ratio for the steel reinforcement

n
= Modular ratio relating the steel reinforcement to the concrete
3.6
118
Figure 3.62 Assumed linear stress distribution in straight line theory
The curvature within the slab, , is then calculated using Equation 3.7
(d k d )
3.7
where
s
d
k
= Curvature in deck slab

= Strain measured in strain gage on RC side bottom transverse rebar
= Distance from the extreme compression fiber to tensile reinforcement
= Effective length factor for compression members (Equation 3.6)
Next, the apparent strain on the RC side, s* , is found at the same distance
along the depth of the deck slab as the strain gages attached to the bottom of the FRP
panels using the formula:
s* = s + (d FRP d s )
3.8
where
s*
= Apparent strain in the RC side deck adjusted to the same depth as gages
attached to FRP panels
= Strain measured in strain gage on RC side bottom transverse rebar
= Curvature in deck slab
119
ds
d FRP
= Distance from extreme compression fiber to top of RC side bottom steel

reinforcement
= Distance from extreme compression fiber to bottom of FRP panel
dFRP
ds
s
s*
Composite side cross-section
Strain distribution
RC side cross-section
Figure 3.63 Illustration of strain distribution and adjustment of RC side strains
Note that Equation 3.8 is relevant for calculating apparent strains in either the
longitudinal or the transverse direction by simply using a different ds value. The
apparent strain at the bottom of the RC side deck can now be more reasonably
compared to the measured strains in the reinforced concrete side.
3.8.5.1 Comparison of Strains in FRP Panel Deck and RC Deck
The load versus strain envelope plot of the bottom longitudinal strain gages
located directly below the RC side and composite side actuators is shown in Figure
3.64. The RC side strain values have been adjusted using the method described in
Section 3.8.5 in order to determine the apparent strains at the depth within the
reinforced concrete side slab that is in line with the FRP panel strain gage.
As observed in Figure 3.64, the two sides maintain comparable strain values up
to a load of approximately 200 kN (45 kip), the load at which the reinforced concrete
120
side load versus strain curve degrades significantly from its original slope due to
cracking of the concrete. Note that the composite side strain gage maintains its initial
slope up until the 400 kN (90 kip) load level, which is over four times the AASHTO
service load. The secondary slope of the load versus strain curve for the composite
side remains approximately linear for the remainder of the test specimen loading to
ultimate capacity and is shallower than the secondary slope for the reinforced concrete
side. The composite side strain values remained lower than the reinforced concrete
side strains for loadings up to approximately 500 kN (112 kip) load level, which is
approximately 5.1 times the AASHTO service load, because the FRP panel side strain
gage reaches a load of nearly double the load taken by the RC side strain gage prior to
significant change in the initial load versus strain response.
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
Load (kN)
DBL17-3* CM3L
CM3L DBL17-3*
500
1000
1500
Strain (microstrains)
2000
2500
Figure 3.64 Load versus strain for bottom longitudinal strain gages
(Gages are located directly below the two actuators. CM3L is on the composite side
on the underside of the FRP panel; DBL17-3 is on the RC side bottom longitudinal
steel reinforcement and therefore is adjusted using Equations 3.7 and 3.8)
121
The adjusted strain in the reinforced concrete side bottom longitudinal strain
gage at the 489 kN (90 kip) load level, which is approximately 5.0 times the AASHTO
service load, was 859.1 microstrains, whereas the strain in the corresponding strain on
the FRP panel longitudinal gage was 665.4 microstrains. At the ultimate capacity load
of 890 kN (200 kip), the strains in the RC side bottom longitudinal strain gage located
directly below the RC side actuator and in the composite longitudinal strain gage
located on the bottom side of the FRP panel directly below the composite side actuator
were 1119.9 microstrains and 1714.5 microstrains respectively.
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
CM3T
DBT10-4*
Load (kN)
CM3T DBT10-4*
500
1000
1500
2000
2500
3000
Figure 3.65 Load versus strain for bottom transverse strain gages
(Gages are located directly below the two actuators. DBT10-4 is on the RC side
bottom transverse steel reinforcement; CM3T is on the composite side underside of
the FRP panel and therefore is adjusted using Equations 3.7 and 3.8)
The load versus strain envelope plots for the bottom transverse strain gages are
shown in Figure 3.65, where the strain for the reinforced concrete side gage is slightly
122
less than the FRP panel side gage from initial loading, up to a load of 140 kN (31 kip),
which is approximately 1.4 times the AASHTO service load. For higher loadings, the
RC side exhibits significantly higher strain values as compared to the composite side.
For example, at the 489 kN (110 kip) load level, which is approximately 5.0 times the
AASHTO service load, the RC side gage (DBT10-4*) has an adjusted strain of 1079.4
microstrains whereas the composite gage (CM3T) has a strain of 578.3 microstrains,
which equates to a 60.5% difference between the two gages.
The strain profile for the top transverse steel reinforcement along the centerline
of the specimen is shown in Figure 3.66. As expected for a top rebar under the given
loading conditions, the strain gages located at the middle of each deck display
compression strains while the gages adjacent to the stems display tensile strains. For
each of the displayed load levels, the reinforced concrete side strain gages exhibit
comparable or higher values as compared to the corresponding strain gages on the
composite side of the test specimen. While Figure 3.66 clearly shows the trend that
the reinforced concrete side is experiencing higher strains, especially at higher load
levels, the individual comparisons of load versus strain plots for pairs of strain gages
are useful in quantifying the difference in response between the two sides.
1000
890 kN
845 kN
756 kN
500
578 kN
0
98 kN
222 kN
400 kN
-500
-1000
-1500
0
610
1220
1830
2440
3050
3660
4270
4880
123
Figure 3.66 Strain profile for top transverse rebar along centerline
124
The load versus strain responses for the top transverse strain gages located
directly below the two actuators (Figure 3.67) illustrates the higher strain values
maintained for the top transverse reinforced concrete side strain gage throughout the
loading as compared to the composite side strain gage. The reinforced concrete side
top transverse strain gage increases nonlinearly up until the 756 kN (170 kip) load
level and then it exhibits discontinuities in the remainder of its strain profile. In
contrast, the composite side top transverse strain gage displays a far more linear
response prior to discontinuities observed at an actuator load of 825 kN (187 kip),
which is approximately 8.4 times the AASHTO service load and is the load at which
initial signs of punching shear were observed.
Load (kN)
1000
DTT10-6 DTT10-3
9x AASHTO 900
service load 800
700
DTT10-3 DTT10-6
600
500
400
300
200
1x AASHTO 100
service load
0
-2000
-1500
-1000
-500
500
Figure 3.67 Load versus strain for top transverse strain gages
(Both gages are located on the top transverse steel reinforcement directly below the
two actuators. DTT10-6 is on the RC side and DTT10-3 is on the composite side)
125
Figure 3.68 shows the load versus strain envelope plots for the top transverse strain
gages which are located 152 mm (6 in) away from either side of the central stem and
are 610 mm (2 ft) away from the centerline of the two deck slabs, in the direction of
the center stem. The two strain gages showed exhibit equivalent strain values up to a
load of 655 kN (155 kip), which is 6.7 times the AASHTO service load. Beyond this
load, the reinforced concrete side increased in strain values more quickly such that the
strain in the RC side gage (DTT10-5), immediately prior to the discontinuities
observed in the load versus strain plots at a load of 829 kN (186 kip) due to the onset
of punching shear, was 2.9 times larger than the strain in the composite side strain
gage (DTT10-4).
DTT10-4 DTT10-5
DTT10-4 DTT10-5
Load (kN)
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
-500
500
1000
1500
2000
Figure 3.68 Load versus strain for top transverse strain gages near central stem
(Both gages are located on the top transverse rebar directly along the test specimen
centerline, each a distance of 610 mm away from the midspan of the decks, towards
the central stem. DTT10-5 is on the RC side and DTT10-4 is on the composite side)
126
The load versus strain envelope plots of the top transverse strain gages located
610 mm (2 ft) away (in the direction of the outer stems) from the centerlines of each
deck slab and 178 mm (7 in) away from the inside edge of the outer stems are shown
in Figure 3.69. For loading of the test specimen up to approximately 400 kN (90 kip),
which is approximately 4.1 times the AASHTO service load, the two strain gages both
display negligible strain values, with the composite side slightly in compression and
the RC side slightly in tension. For loads greater than 400 kN (90 kip), the RC side
gage increases in strain values with increasing load significantly more than the
composite side gage.
DTT10-2
DTT10-7
DTT10-2
DTT10-7
Load (kN)
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
-500
500
1000
1500
2000
Figure 3.69 Load versus strain for top transverse strain gages near outer stems
the outer stems. DTT10-7 is on the RC side and DTT10-2 is on the composite side)
127
The FRP panel side strain gage remains slightly in compression until a load of
approximately 700 kN (157 kip) and then begins to exhibit tensile strains, which
increase with load at approximately the same rate as the RC side strain gage. At the
ultimate capacity load of 890 kN (200 kip), the composite side strain value is 400.9
microstrains while the RC side strain value is over twice as large at 869.0
microstrains.
3.8.5.2 Strain Profiles along the Longitudinal Direction of the Deck

It is important to determine the degree to which the strains due to live loading
distribute along the deck in both directions. The strain profiles along the longitudinal
directions for the centerlines of the composite side and reinforced concrete side deck
slabs are shown in Figure 3.70 and Figure 3.71 respectively. Note that as discussed in
Section 3.8.5, the position of the strain gages within the composite side and RC side
deck of the test specimen with respect to the depth of the slab will have an effect on
the direct comparability of the absolute magnitude of the strains. Since the FRP side
does not have a bottom layer of rebar, the strain gages for the bottom layer on the
composite side are attached the underside of the FRP panels, which is located a
distance of approximately 57 mm (2.25 in) lower within the deck slab than the gages
attached to the bottom longitudinal rebar on the reinforced concrete side.
Regardless of the magnitude of the strain values, the strain profiles of two deck
slabs can be compared in terms of profile geometry and loads at which major changes
occur. Both sides maintain negligible longitudinal strain values of less than 100
128
microstrains for the central strain gage up to the load level of 222 kN (50 kips), which
is 2.3 times the AASHTO service load. At the 400 kN (70 kip) load level, which is
over four times the AASHTO service load, the reinforced concrete side center strain
gage exhibits a significantly greater strain while the composite side maintains a low
strain value. By the 578 kN (90 kip) load level, which is equivalent to 5.9 times the
AASHTO service load, the composite exhibits a significantly higher strain value from
its initial strain. For further information on the center longitudinal strains, see page
120 of Section 3.8.5.1, which contains a more detailed discussion.
The composite side and reinforced concrete side deck slabs shown in Figure
3.70 and Figure 3.71 maintain very comparably shaped profiles throughout most of the
loading range, which both display large tensile values in the strain gage directly below
the actuator, compressive strains in the strain gage approximately 610 mm (2 ft) away
from the actuator and smaller compressive strains in the strain gages farthest away
from the actuator. The switch in strain values from tension to compression along the
strain profile indicates that the point of inflection for both slabs in the longitudinal
direction is less than 610 mm (2 ft) longitudinally away from the centroid of the
actuators. Since the intersection between the FRP panels is located 610 mm (2 ft)
longitudinally away from the centerline of the actuators which is very near the point of
inflection for both of the deck slabs it is difficult to reliably determine the
effectiveness of stress transfer between the panels from the current experiment.
Therefore, it is recommended that the effectiveness of stress transfer between panels
be examined in greater detail through future experimental studies.
129
3000
2500
2000
1500
1000
500
0
890 kN
845kN
756 kN
+
"0"
578 kN
400 kN
222 kN
98 kN
-500
-1000
-1830
-1220 -610
0
610
1220
1830
Figure 3.70 Strain profile for gages attached to bottom of FRP panel along line 3
(Line of gages is located directly along the composite side centerline)
2000
1500
1000
500
0
-500
-1830
"0"
890 kN
845 kN
756 kN
578 kN
400 kN
222 kN
98 kN
-1220
-610
0
610
1220
1830
Figure 3.71 Strain profile for bottom longitudinal rebar along line 17
(Rebar is offset from RC side centerline by 140 mm towards the center stem)
The shift from compression to tension values at the 890 kN (200 kip) load
level for the bottom longitudinal RC side strain gage shown in Figure 3.71, located
130
610 mm (2 ft) away from the middle of the specimen, is presumably due to a
discontinuity which was observed during the 845 kN (190 kip) loading phase at which
the initial signs of punching shear were observed on the FRP panel side deck slab.
The strain profile for gages attached to the top longitudinal steel reinforcement
along the centerline of the composite is shown in Figure 3.72. This plot shows the
same profile geometry as the bottom longitudinal profiles for both deck slabs, with
tensile strains which become appreciable at the 400 kN (70 kip) load level and a point
of inflection at a location prior to the strain gage located 610 mm (2 ft) away from the
centerline of the test specimen. Note, however, that the top longitudinal strain profile
maintains the same signs as the bottom longitudinal strain profile such that the center
strain gage is in tension and the two farther away strain gages are in compression,
which is indicative of more localized deformations in the region.
1250
1000
750
500
250
0
"0"
890 kN
845 kN
756 kN
578 kN
400 kN
222 kN
98 kN
-250
-500
-1830
-1220
-610
0
610
1220
1830
Figure 3.72 Strain profile for top longitudinal rebar along line 8
(Rebar is offset from composite side centerline by 140 mm towards center stem)
131
3.8.6
Strain Progression and Maximum Strain Levels
At the 1x AASHTO service load of 98 kN (22 kip), all strains in the test
specimen were less than 100 microstrains, with the highest strains measured in the
strain gages directly below the two actuators. At the 489 kN (110 kip) load level,
which is equivalent to approximately five times the AASHTO service load, the highest
strain value of 1446.8 microstrains was recorded from the bottom transverse RC side
strain gage at midspan located 203 mm (8 in) longitudinally away from the centerline
of the RC side actuator.
The bottom transverse strain gage directly below the
reinforced concrete side and composite side actuators recorded strain values of 762
microstrains and 578 microstrains respectively at this load level. The second highest
strain value of -794 microstrains was located on the same bottom transverse bar as the
highest recorded strain value at this load level on the segment which extends onto the
composite side, at a distance of 152 mm (6 in) away from the edge of the center stem.
The bottom longitudinal strain gage readings on the strain gages located directly
below the reinforced concrete and composite actuators, were 545 and 665 respectively
for this load level.
At the 578 kN (130 kip) load level, equivalent to nearly six times the
AASHTO service load, a few of the strain gages in the deck measured sudden shifts in
strain values.
Most notably, the bottom RC side transverse strain gage which
displayed the maximum value during the 489 kN (110 kip) load level measured a
sudden jump from 1252 microstrains to a strain value of 1919 microstrains, which was
the highest measured strain value in the test specimen at this load level. The other
notable strain discontinuity was observed in the composite side transverse strain gage
132
attached to the bottom of the FRP, located along the test specimen centerline and 152
mm (6 in) from the edge of the center stem. The discontinuities in the strain profile
for this strain gage are displayed in Figure 3.78 on page 142.
At the 756 kN (150 kip) load level, equivalent to approximately 7.7 times the
AASHTO service load, strain discontinuities occurred throughout the reinforced
concrete side, primarily within the strain gages attached to the bottom layer of steel
reinforcement. Conversely, the composite side only displayed sudden shifts in strain
values in the two transverse strain gages attached to the bottom of the center FRP
panel, along the test specimen centerline, located 152 mm (6 in) and 178 mm (7 in)
from the edge of the center and outer stems respectively. The maximum strain in the
deck slab was 2150 microstrains for the bottom transverse RC side strain gage at
midspan located 203 mm (8 in) longitudinally away from the centerline of the RC side
actuator. The strains for the bottom transverse RC side and composite side center
gages at this load level were 1422 microstrains and 719 microstrains respectively
whereas the bottom longitudinal strains on the RC side and composite side were 860
microstrains and 1743 microstrains respectively.
At the 845 kN (190 kN) load level, equivalent to 8.6 times the AASHTO
service load, significant strain discontinuities occur throughout the reinforced concrete
side deck slab. The most notable strain discontinuities occurred in the RC side bottom
transverse rebar strain gages located a distance of 152 mm (6 in) away from the edge
of the center stem, with the strain gage located along the centerline of the test
specimen abruptly changed from -302 microstrains to -949 microstrains whereas the
strain gage located 610 mm (2ft) longitudinally away from the centerline of the
133
actuators changed from 200 microstrains to 5767 microstrains. Similar abrupt changes
in strain were not observed on the composite side.
At ultimate capacity, which occurred at a load of 890 kN (200 kip), equivalent
to approximately nine times the AASHTO service load, the highest overall strains in
the test specimen were observed in the RC side bottom transverse rebar gages located
below the RC side actuator and adjacent to the edges of the center and outer stems,
which read 5796 microstrains and 2975 microstrains respectively.
The bottom
transverse and longitudinal strains measured on the strain gages attached to the bottom
of the FRP panel, directly below the composite side actuator were 920 microstrains
and 2442 microstrains respectively and the corresponding bottom transverse and
longitudinal strains measured on the strain gages attached to the steel reinforcement
directly below the RC side actuator were 1930 microstrains and 1120 microstrains
respectively. The largest compressive strain in the deck slab of -1676 microstrains
was measured in the bottom transverse rebar located along the test specimen centerline
and positioned on the composite side a distance of 152 mm (6 in) away from the edge
of the center stem.
3.8.7
Unloading Response
3.8.7.1 Loading versus unloading stiffness comparisons

A comparative evaluation of the degradation of the slope of the load-deflection
response of the slabs during the unloading portions of the test was obtained using the
effective stiffness, which was defined earlier in Equation 3.1 on page 87 of Section
134
3.8.3. The effective stiffness ratio was then computed using the same formulation
defined earlier in Equation 3.2 and the equation is repeated here to highlight the use of
a different load level for the baseline:
Effective Stiffness Ratio =

3.2
Note that the baseline for the unloading portions of the test is the 222 kN (49.9
kip) load level rather than the 98 kN load level used for the loading portions of the
curve because the 222 kN (50 kip) load level is the first level at which a single
unloading cycle is performed. The effective stiffness ratio values for the FRP side
slab and the reinforced concrete side slab during unloading sections of the test are
presented in Table 3.12 and Table 3.13 respectively.
At each load level, the reinforced concrete side exhibits a lower effective
stiffness ratio as compared to the composite side. This indicates that the slope from
the unloading portion of the load versus deflection plot for the reinforced concrete side
degrades more rapidly than the equivalent values for the composite side. The effective
stiffness of the reinforced concrete side also degrades to a greater extent overall, as
observed by the 25 % lower effective stiffness ratio of the reinforced concrete side
during the final unloading cycle at 756 kN (170 kip) prior to reaching the ultimate
capacity of the test specimen.
135
Table 3.12 Effective stiffness ratio of FRP side slab from linear potentiometer data
Table 3.13 Effective stiffness ratio of RC side slab from linear potentiometer data
A comparison of the effective stiffness ratios for the loading and unloading
portions of the test, shown in Figure 3.73, indicates that the reinforced concrete side
exhibits significantly lower effective stiffness ratios as compared to the composite side
for both the loading and unloading portions of the curves. Note that the effective
stiffness ratios for the loading portions of the test were recalculated using the baseline
of 222 kN (50kip) to enable the values to be accurately compared to the effective
stiffness ratios for the unloading portions of the test.
136
1.2
composite sideloading portion
1.0
0.8
composite sideunloading portion
0.6
RC sideloading portion
0.4
0.2
RC sideunloading portion
0.0
222
400
578
Load Level (kN)
756
Figure 3.73 Stiffness ratios of the center of the two slabs found from deflection data
(loading portion data was recalculated with 222kN instead of 98 kN as the baseline)
3.8.7.2 Residual deflections and strains

The presence of residual deflections and residual strains observed in the test
specimen deck slab throughout the loading protocol are an indication of irreversible
damage which has occurred. Throughout the loading steps, the residual deflection
accumulated for each load level at the midspan of each slab was calculated and the
results are displayed in Table 3.14. Note that the reinforced concrete side exhibited
higher residual deflection values for each load level as compared to the composite side
values. The total residual deflection accumulated with loading up to the 756 kN (170
kip) load level was 1.21 mm (0.048 in) for the composite side as compared to a 62.9%
greater total residual deflection for the reinforced concrete side of 1.97 mm (0.078 in).
137
Table 3.14 Residual deflections of deck slabs (from central linear potentiometers)
A similar phenomenon is observed in the strain data, in which the reinforced

concrete side exhibits significant residual strains for the many of its deck slab gages,
while the composite side exhibits significantly smaller and often negligible residual
strains for its deck strain gages.
For illustration purposes, Figure 3.74 shows a
representative load versus strain plot, which includes the unloading portions of the
curve. The reinforced concrete side strain gage exhibits significantly greater residual
strains than the composite side strain gage and a similar trend is observed throughout
the test specimen.
The considerably greater residual deflections and strains observed on the
composite side are a good indication of increased damage which has occurred
throughout the loading in the examined regions. The smaller residual strains and
deflections on the composite side can be partially attributed to the FRP panels, which
act mitigate the impact of damage in the deck by catenary action in which the FRP
138
panels act compositely with the concrete and resists crack development in the
concrete.
DTT10-2
DTT10-7
DTT10-2
Load (kN)
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
-500
500
1000
DTT10-7
1500
2000
Figure 3.74 Representative load versus strain plot to illustrate difference in residual
strains between composite side and RC side strain gages.
(Specific plot shown contains gages on top transverse rebar along centerline, located
178 mm away from of the inner edge of the outer stems and 610 mm away from the
centerline of the slabs)
3.8.8
Separation between FRP Panels and Concrete
During the static loading of the test specimen, examination of specific strain
and deflection data was used in order to detect the presence of a separation between
the FRP panels and the concrete. A select number of linear potentiometers were
placed above and below the composite deck in the same location such that their
deflections could be compared. The difference in deflection values obtained from
pairs of these linear potentiometers was used to obtain information regarding the level
of separation that had occurred.
A differential between the top and bottom pot
139
readings indicates that the concrete and composite are no longer acting together as a
connected unit and are now deflecting independent of each other.
The load versus deflection envelope plots of the top and bottom pair of linear
potentiometers located along the transverse centerline of the test specimen on either
side of the actuator on the FRP panel deck are shown in Figure 3.75 and Figure 3.76.
Based on the observed fluctuations in measured deflections for top and bottom linear
potentiometer pairs on the reinforced concrete side deck, a tolerance level of 0.16 mm
(0.006 in) was established as the threshold value for determining that a difference
between the top and bottom deflections was significant.
For the pair of linear
potentiometers located near the outer stem shown in Figure 3.75, a difference of 0.17
mm (0.007 in) between upper and lower deflection values was observed at a load of
673 kN (151 kip), which is approximately 6.9 times AASHTO service load, which
increased to 1.09 mm (0.043 in) at ultimate capacity.
For the pair of linear
potentiometers located near the center stem shown in Figure 3.76, the upper and lower
deflection values were separated by 0.17 mm (0.007 in) at a load of 489 kN (109 kip)
and the difference increased to 3.12 mm (0.123 in) at ultimate capacity.
Another method for detection of the separation between the concrete and the
FRP panels compares strain gauges which are attached to the composite panel, the
bottom rebar which protrudes onto the composite panel in order to satisfy the
development length required by code, and the top rebar layer which is continuous
throughout the specimen. The vertical difference between the strain gauge on the
bottom of the FRP plate and the strain gauge on the bottom rebar is approximately
50mm (2 in) and so similar strain values are expected, with the strain gage on the
140
bottom of the FRP plate reading slightly higher strains than the strain gage on the
bottom rebar.
1000
900
800
700
600
500
400
300
200
100
0
Load (KN)
Top Bottom
3
4
5
Deflection (mm)
Figure 3.75 Load versus deflection plots for detection of FRP panel separation
(located along test specimen centerline and 610 mm away from the composite side
centerline, 178 mm from the inner edge of the outer stem)
1000
900
800
700
600
500
400
300
200
100
0
Bottom
Load (KN)
Top
3
4
5
Deflection (mm)
Figure 3.76 Load versus deflection plots for detection of FRP panel separation
(located along test specimen centerline and 610 mm away from the composite side
centerline, 152 mm from the composite side edge of the center stem)
141
As shown in Figure 3.77, the strain values for the bottom rebar and the
composite panel are very similar until a load of approximately 700 kN (157 kip) then
the strains diverge suddenly, whereas Figure 3.78 shows a significant difference in
strain values initiating at a load of 500kN (112 kip). The discontinuity in the strain
profile between the FRP panel and the bottom rebar indicate that the FRP panel had
separated from the concrete. Note that the site of initiation for the debond between the
concrete and the composite was adjacent to the support rather than in the center of the
FRP panel, due to the high shear present in this region.
Based on the two detection methods described above, sections of the center
composite panel exhibited initial signs of separation at approximately the 489 kN (110
kip) load level, which is equivalent to 5.0 times the AASHTO service load. Note that
the damage initiation first occurred on the side of the FRP panel adjacent to the center
support because the loading in this region was more severe than in the region adjacent
to the exterior support. Comparisons based on linear potentiometers and tap tests at
different locations throughout the test specimen indicated that the concrete to
composite bond for the two outer panels remained fully intact throughout the loading
levels.
142
Load (kN)
1000
CM1T DBT10-1 DTT10-2
9x AASHTO 900
service load 800
700
600
Cross section of deck
Plan view
500
DTT10-2
400
300
DBT10-1
200
CM1T
1x AASHTO 100
service load
0
-2000 -1500 -1000
-500
0
500
1000
specimen centerline at the same location and at different depths within the deck
Load (kN)
1000
DBT10-2
DTT10-4 CM5T
9x AASHTO 900
service load 800
700
600
Plan view
500
400 DTT10-4
300
DBT10-2
200
CM5T
1x AASHTO 100
service load
0
-2000 -1500 -1000
-500
0
500
1000
143
3.9
LIMIT STATES
In general, a structure must be designed such that the requirements for both
serviceability and strength limit states are satisfied. The serviceability limit states
relate to the comfort of the occupants in terms of limiting excessive deflections,
cracking and vibrations. The strength limit states relate to the structural collapse of all
or part of the structure in terms of stability or buckling, rupture of critical parts of the
structure, the formation of plastic mechanisms in ductile members [154]. These
strength limit states are checked for different loading conditions such as flexure, shear,
bond strength and buckling. In addition to serviceability and strength, Section 5.5 of
the AASHTO LRFD Bridge Design Manual includes provisions for fatigue and
fracture limit states as well as an extreme even limit state [91]. Fatigue and fracture
limit states are intended to prevent the failure of a structure through the progression of
cracks and other defects. The extreme event limit state is intended to resist structural
collapse due to extreme events such as vehicle collisions, floods, or earthquakes.
Certain limit states such as deflection and vibration limitations are material
independent, whereas of other limit states are different depending on the specific
building materials chosen. A brief overview of limit states for reinforced concrete
structures and hybrid concrete-FRP structures is included below.
For both steel reinforced concrete structures and hybrid reinforced concreteFRP structures, limits are placed on the allowable stresses in the steel reinforcement to
avoid yielding of the steel reinforcement and inelastic deformations under service
loads as described in AASHTO, section 5.7.3.4 [91] and ACI 440, section 9.4 [148]
144
respectively. At service load, the AASHTO defined the maximum deck deformation
due to local dishing at wheel loads using the following criteria:
L/800 for decks with no pedestrian traffic

L/1000 for decks with limited pedestrian traffic
L/1200 for decks with significant pedestrian traffic
where
L
= Span length of the deck measured from center-to-center of supports
To calculate the immediate deflections due to service loads for concrete or hybrid
concrete-FRP, the effective moment of inertia, I e , as given in ACI 318-05 Section
9.5.2.3 [155], is used:
M
I e = cr
Ma
3
M
I g + 1 cr
M a
I cr I g
3.9
in which
M cr = f r
Ig
3.10
yt
where
M cr = Cracking moment (ksi)
Ma
= Maximum moment in a component at the stage for which deformation is
computed (kip-in)
I cr
= Moment of inertia of cracked section transformed to concrete (in4)
Ig
= Gross moment of inertia (in4)
fr
yt
= Modulus of rupture of concrete (ksi)

= Distance from the neutral axis to the extreme tensile fiber (in)
The modulus of rupture for concrete, f r , is defined by AASHTO as
f r = 0.24 f c'
for normal weight concrete
f r = 0.20 f c'
for sand-lightweight concrete
f r = 0.17 f c'
for all-lightweight concrete
145
where
f c'
= Specified compressive strength of concrete at 28 days (ksi)
For strength limit states, the factored resistance values are determined through
multiplying the calculated nominal resistance values by their corresponding resistance
factors, . Some of the representative resistance factors used for conventionally
constructed steel reinforced concrete structures are as follows:
= 0 .9
= 0.9
= 0 .7
for flexure and tension of reinforced concrete

for shear and torsion of normal weight concrete
for bearing on concrete
In addition to the above resistance factors, FRP-strengthened sections use additional

strength reduction factors to account for the effects of different exposure conditions,
larger material variability, and potentially lower ductility. For example, the design
ultimate tensile strength of the FRP reinforcement, f fu , is defined by ACI 440.2R as:
f fu = C E f fu*
3.11
where
CE
= Environmental reduction factor defined in ACI440.2R Table 8.1. Values are
between 0.50 and 0.95 based on fiber and resin types and exposure
conditions
*
f fu
= Ultimate tensile strength of the FRP as reported by the manufacturer
A reinforced concrete section in flexure is checked for the strength limit states
of crushing of the concrete in compression before yielding of the reinforcing steel as
well as yielding of the steel in tension followed by concrete crushing. Additional
146
failure modes considered for an FRP-strengthened section as described in ACI 440.2R
Section 9.2.1 are yielding of the steel in tension followed by rupture of the FRP
laminate, shear/tension delamination of the concrete cover and debonding of the FRP
from the concrete substrate [148].
3.10 DESIGN THRESHOLDS
While experimental results indicate that the composite panel reinforced bridge
deck had an ultimate capacity of over nine times the AASHTO service load and
deflected less than the reinforced concrete side deck throughout the loading, it is
necessary to determine the loads which the FRP panel system can be safely support
without undergoing significant irreversible damage.
Since the initial signs of
separation between the FRP panels and the concrete occurred at the 489 kN (110 kip)
load level, this load will be taken as the maximum threshold for the FRP panel system.
This serviceability limit state as well as the experimentally determined strength limit
states for punching shear and ultimate capacity of the structure are shown in Table
3.15 in terms of the code defined loads of Service I, Strength I [91], and the Permit
Truck [152] load.
Based on the information regarding the observed limit states for the system and
the corresponding safety factors for different permit loadings, an acceptable design
load for the examined FRP panel system would be 360 kN (82 kip). This would allow
for a safety factor of 2.0 for the CALTRANS permit truck and a safety factor of 3.7
and 2.1 for the AASHTO Service I and Strength I load levels respectively. Note that
147
this design threshold is directly tied to the geometry of the specimen and substantial
changes in dimensions could cause significant changes in design threshold values.
Table 3.15 Safety factors for different limit states in terms of permit loadings
3.11 CYCLIC LOADING OF TEST SPECIMEN

3.11.1 Loading Protocol
After the ultimate capacity of the test specimen was reached, a cyclic phase
was implemented.
Since the test specimen had already been loaded to ultimate
capacity, this loading phase was not an accurate representation of response under
actual cyclic loading. Instead, the limited cyclic loading of the test specimen provided
useful information regarding further accumulation of damage within the specimen. If
an unforeseen event applies loads significantly larger than typical to a bridge structure,
it is important to determine whether the structure can still safely support normal traffic
148
loads. If the systems post failure response to cyclic loading is very stable, the bridge
can be left in service while preparations are made to repair the damage caused by the
excessive unforeseen loading to the structure.
The test specimen was loaded for 1,000 cycles from the minimum threshold load of
4.5 kN (1 kip) to 1x AASHTO fatigue service load, which is defined as 84 kN (18.9
kip), and cycled to 2x the fatigue service load for another 1,000 cycles as summarized
in Table 3.16. For cyclic loading, the fatigue service load of 84 kN (18.9 kip) was
defined by the SERVICE I level in which the HS20 wheel load of 73 kN (16.4 kip)
with an additional AASHTO defined 15% dynamic allowance for the fatigue limit
state was applied.
Table 3.16 Cyclic loading protocol for test specimen #1
Center deflections were chosen as the primary indicator by which to make

comparisons regarding stability and progression of damage. Note that the deflection
values were biased using the original bias values from the start of static testing in
order to establish the cumulative deflections present in the deck. Since the stability of
149
the system during the cyclic loading was the focus of this loading phase, changes in
deflection and strain values over the course of the cyclic loading rather than their
absolute magnitudes are of primary concern.
3.11.2 Changes in Center Deflections during Cyclic Loading
The center deflections during the course of cyclic phases I and II are shown in
Table 3.17 and Table 3.18 respectively. As observed in both tables, the changes in
center deflections for the FRP panel side were less than the experimentally determined
threshold of 0.17 mm (0.007 in) for establishing whether a deflection measurement
was statistically significant. Conversely, the reinforced concrete side was slightly less
than the deflection threshold for cyclic phase I and was slightly over the deflection
threshold for cyclic phase II.
So, the changes in the center deflections for the
reinforced concrete side of 0.20 mm (0.008 in) over the course of the cyclic loading
phase II amounts to a 5.7% increase in deflection values. While this deflection
increase is small, it is still considered statistically significant because the values are
above the defined threshold of 0.17 mm (0.007 in), unlike the changes in deflection for
the FRP panel side throughout both cyclic phases that fall well below the minimum
deflection threshold.
150
Table 3.17 Center deflections during cyclic loading phase I (84 kN target load)
Table 3.18 Center deflections during cyclic loading phase II (168 kN target load)
The effective stiffness ratios for the two decks, shown in Table 3.19 and Table
3.20, were calculated for both cyclic loading phases using the adaptive secant stiffness
method described on page 87 in Section 3.8.3. Since the load acting on the system
remains constant throughout the cyclic phase, any change in the stiffness ratio is
attributed to a change in deflections.
During cyclic phase I, the effective stiffness
ratios remained constant for both slabs as shown in Table 3.19 and Table 3.20. This is
an indication that negligible structural degradation has occurred in either slab during
cyclic loading phase I of this test specimen.
151
Table 3.19 Effective stiffness ratios for FRP panel side during cyclic loading phase I
Table 3.20 Effective stiffness ratios for RC side during cyclic loading phase I
The progression of the effective stiffness ratios for both sides over the course
of cyclic phase II is shown in Figure 3.79. Table 3.21 shows the effective stiffness
ratios for the FRP panel side dropped by 12% during the first 250 cycles and dropped
by another 5% during the next 750 cycles while Table 3.22 shows the reinforced
152
concrete side effective stiffness ratios remained constant with only 5% change
throughout the loading phase.
Due to the relatively small loads applied to the system during the cyclic phases
(the loading at cyclic phase II is less than five times smaller than the loads at the
ultimate capacity from the static loading phase) the calculations of effective stiffness
are heavily affected by minor changes in deflection values. So, the increase in overall
deflections for the composite side of 0.26 mm (0.010 in) from the start to the end of
cyclic phase II, which corresponded to a 19% decrease in the effective stiffness ratio,
was only 1.5 times the experimentally determined deflection threshold of 0.17 mm
(0.007 in).
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
Composite side
250
500
750
RC side
1000
1250
Cycle #
Figure 3.79 Effective stiffness versus cycle number for phase II of cyclic loading
153
Table 3.21 Effective stiffness ratios for FRP panel side during cyclic loading phase II
Table 3.22 Effective stiffness ratios for RC side during cyclic loading phase II
3.11.3 Changes in Strains during Cyclic Loading
The progressions of strains during the cyclic loading phases for strain gages
located in the deck slab directly below the actuators are presented in Table 3.23 and
Table 3.24. Note that all gages were biased using the original strain gage values from
154
the start of the static testing in order to record the complete history of the strain
imposed on each gage due to all previous loading. Also, all data shown in the tables
below were taken with the specimen at the full target load for a given cyclic phase. A
number of the important strain gages, including the bottom longitudinal gage located
directly below the RC side actuator, were permanently damaged when the specimen
was loaded to ultimate capacity. However, for most of the strain gages that continued
to function properly only minor progressions in strain values during the cyclic loading
phases were observed.
For the strain gages located directly below the FRP side actuator which are
shown in Table 3.23, the strain values changed by less than 15 microstrains or 7%
over the course of the first cyclic phase and changed by less than 20 microstrains or
5% over the course of the second cyclic phase. The top (DTT10-6) and bottom
(DBT10-4) transverse RC side strain gages located directly below the RC side
actuator, shown in Table 3.24, both exhibit comparable progressions in strain to the
corresponding gages on the FRP panel side (DTT10-3 and CM3T) over the course of
cyclic loading phase I. However, during the cyclic loading phase II, the top and
bottom transverse RC side gages shifted by 50.0 and 38.5 microstrains respectively,
whereas the FRP panel gages shifted by only 0.6 and 7.3 microstrains respectively.
Table 3.23 Strain progresssion for composite side gages under target loadslocated directly below the composite side actuator
(See Section 3.6.2 for detailed information on precise location of each strain gage)
155
DESIGNATIONS
CYCLIC LOADING
PHASE II
CYCLIC LOADING
PHASE I
POSITION
Table 3.24 Strain progresssion for reinforced concrete side gages under target loadslocated directly below the RC side actuator
STRAIN GAGE NAME
Location (All on FRP panel side)
Attached to rebar or FRP
Near top or bottom of deck
Oriented longitudinal or transverse
DBT10-4
Below actuator
Rebar
Bottom
Transverse
DTT10-6
Below actuator
Rebar
Top
Transverse
Strain at Cycle # 1 ()
Change in strains ()
Change in strains (%)
484.2
486.5
487.1
488.3
497.5
13.2
2.7%
-853.9
-899.1
-909.8
-845.5
-839.6
14.3
-1.7%
655.0
665.9
692.3
698.1
693.5
-634.14
-561.50
-563.89
-579.36
-584.13
Change in strains ()
Change in strains (%)
38.5
5.9%
50.0
7.9%
156
157
3.11.4 Progression of Separation between FRP Panel and Concrete
The FRP panel side strain gages located adjacent to the two FRP side stems
shown in Table 3.25 are the gages used during static testing as indicators of separation
between the concrete and composite panels. Unfortunately, the strain gages attached
to the bottom transverse reinforcement at the same locations were both broken and so
a complete discussion of strain discontinuities at these locations is not possible.
However, because the most severe separation occurred in these regions during the
static loading to ultimate capacity, significantly larger progressions in strain over the
course of the cyclic loading are observed, as shown in Table 3.25, as compared to the
strain gages directly below the actuators. Also note that the strains in the gages
adjacent to the center stem are significantly more stable than those located adjacent to
the outer stem.
The progression of the separation between the concrete and the composite at
the same locations examined in Section 3.8.8 is summarized in Table 3.26. The both
pairs of linear potentiometers display relatively constant results during the first cyclic
phase. During the second cyclic phase, the linear potentiometers located near the
central stem continue to display very similar separation values whereas the linear
potentiometer pair near the outer stem exhibits a 0.19 mm (0.008 in) decrease in the
overall separation distance between the concrete and the FRP panel.
One possible explanation for the difference in the stability levels of the gages
and separation readings for two locations is that the FRP panel had fully separated
from the concrete in the region near the center stem and so exhibited a stable response
during the cyclic loading.
Table 3.25 Strain progresssion for transverse composite side gages under target loadslocated along the centerline, near the stems
158
Table 3.26 Separation progresssion for linear potentiometers located along the test speciment centerline, near the stems
(See Section 3.6.1 for detailed information on precise location of each linear potentiometer)
159
160
Conversely, the region of the FRP panel near the outer stem may display a less
stable response as the separation between the concrete and the FRP panel propagates
during the cyclic loading phases.
3.12 SUMMARY
The system level structural response of the FRP panel structural formwork
system integrated within a two cell concrete box girder configuration was
characterized through static and limited cyclic testing.
Direct performance
comparisons between the FRP structural formwork system and a reinforced concrete
system were addressed. Aspects of the FRP panel reinforced systems response,
including the overall response, the response between the panels, the strain
development within the structure, changes in stiffness, and the progression of damage
within the system were evaluated and discussed.
As a result of a direct performance comparison to the reinforced concrete deck
slab, the FRP panel system was seen to perform either comparable or superior in
nearly all aspects examined. The FRP panel side experienced less degradation of the
effective stiffness, lower strain values throughout most of the deck slab, less residual
deflections and lower center deflections throughout the loading range, especially at
higher load levels. Due to the presence of the stiffeners on the FRP panels, the FRP
panel system had a significantly more dispersed deflection distribution in the
transverse direction and a slightly more localized deflection distribution in the
longitudinal direction as compared to the deflection distributions for the reinforced
concrete deck slab.
161
The overall response of the FRP panel system was observed to surpass
specified load requirements by a substantial margin, with the systems ultimate load
carrying capacity of 890 kN (200 kip) equivalent to 9.1 times the AASHTO service
load and 4.9 times the CALTRANS permit truck load [152]. Signs of separation
between the concrete and the FRP panels were first observed at the 489 kN (110 kip)
load level, which was 5.0 times the AASHTO service load and 2.7 times the
CALTRANS permit truck load. By using the load level at which initial signs of
separation between the FRP panels and the concrete were observed as the critical state,
an appropriate design load was determined for the FRP panel system.
Initial signs of the punching shear failure mode for the FRP panel side were
observed at a load of 825 kN (186 kip), which is equivalent to 8.4 times the AASHTO
service load. Since the punching shear response occurred at a load substantially above
the specified loading requirements, it is unlikely that this potential failure mode would
be critical at any point during the service life of the system. The punching shear
failure was non-catastrophic and the systems post failure response to applied cyclic
loading was very stable. The stability of the response indicates that even after partial
loss of composite action between the FRP panels and the concrete ocurred, the bridge
could remain open to traffic, albeit with slightly higher deflections during traffic
loading. The ability to allow the damaged bridge to remain open to normal vehicular
traffic is especially crucial in locations where road closures can cause significant
detrimental impact on the surrounding community.
4
4.1
REPAIR OF THE FRP PANEL STRUCTURAL FORMWORK SYSTEM

INTRODUCTION
Following the completion of the cyclic phase for the first test specimen, an
epoxy injection technique, which utilized a low viscosity resin, was developed and
implemented to repair the damage that was incurred during structural testing.
The
separation between the concrete and the FRP panels as well as damage due to
punching shear were addressed by this repair technique. The repair was analogous to
resin injection techniques performed for reinforced concrete bridges as described in
Section 2.3. The presence of the FRP structural formwork panels had no significant
negative impacts on the ease of the repair and instead provided the added convenience
of eliminating the need to provide material below the underside of the deck to catch
excess resin that had flowed through the full thickness of the deck.
The goal for this phase of the research was to assess the effectiveness of repair
efforts at reinstating proper load transfer and bond between the concrete and the
composite panels in terms of field applicability and impact on structural performance.
Aspects of the response of the FRP panel reinforced system including the overall
response, the response between the panels, changes in stiffness, and the progression of
damage within the system were evaluated and discussed. The repaired specimen was
loaded statically up to 978kN (220 kip), which was 88 kN (20 kip) beyond the
ultimate capacity of the original test specimen and the system did not reach its
ultimate capacity. The repaired specimen was then loaded for 500 cycles at 2x the
AASHTO fatigue factored service load and then reloaded statically up to the 978kN
162
163
(220 kip) load level. Since the primary purpose of testing the repaired specimen was
to compare the structural response of the repaired FRP panel side to the original test
results, only information relevant to the FRP panel side deck are included in the
discussion of experimental results.
4.2
DESIGN AND IMPLEMENTATION OF REPAIR
The extent of the separation between the concrete and the composite was
established by performing a tap test on the underside of the composite side deck. The
separated region of the composite side deck is mapped in red, as shown in Figure 4.1.
Figure 4.1 Region in which the FRP panels had separated from the concrete
Note that the separation between the concrete and composite is constrained to
the region between stiffeners, such that the majority of the center FRP panel has
164
disconnected from the concrete with the separation arrested at the stiffener walls. The
outer FRP panel shown in yellow remained fully intact, whereas the outer FRP panel
shown in orange was only separated from the concrete along its length for the 127 mm
(5 in) wide section which was adjacent to the center panel. This
phenomenon
highlights a feature of the FRP structural panel system; the debonded regions are
constrained between stiffeners thereby slowing down the progression of damage
throughout the system.
The repair technique was implemented by drilling multiple small holes in the
top of the composite side deck concrete in a periodic arrangement of 19 mm (.75 in)
diameter holes for resin injection ports and 13 mm (0.5 in) diameter holes for vent
holes as shown in Figure 4.2 by the blue and red circles respectively. This design
allowed the resin to flow within each trough created by the FRP panel stiffeners and
act to reinstate load transfer between the concrete and the FRP panels. Note that the
grey rectangle shown in Figure 4.2 represents the footprint of the elastomeric bearing
pad for the FRP side actuator, which was left in place on the specimen during the
repair. Plastic fittings were bonded into each hole using a two part epoxy system and
a low viscosity resin system was pumped into the holes using a Masterflex peristaltic
pump and the setup shown in Figure 4.3. The resin was pumped into each injection
port until the tubing which transported the resin began to swell up with excess resin
due to built up backpressure. When this occurred, the tubing was sealed off using a
tubing clamp, the tubing was cut and the system was moved to pump resin into the
next available injection port.
165
Figure 4.2 Detail of hole pattern arrangement for repair of FRP side deck
(Only the FRP panel side deck and not the RC side deck shown. Region encompassed
by red line indicates where separation between FRP panel and conrete has occurred)
166
Figure 4.3 Injection port in top of deck and setup used to inject resin into deck
The Sikadur 55 SLV resin system from Sika Corporation used for the resin
injection has a viscosity of approximately 105 centipoise and was mixed in a 2:1 ratio
by volume of Component A to Component B [156]. The ambient temperature at
the time of the repair was 16 C (61 F). Selected manufacturer provided properties
for the system are summarized in Table 4.1. The resin system used in the repair of the
FRP panel side deck was allowed to cure for 19 days prior to testing.
Table 4.1 Properties of Sikadur 55 SLV resin system [156]
167
INSTRUMENTATION
All instrumentation was disconnected briefly from the signal conditioning and
data acquisition systems so that the instrumentation could be to temporarily moved to
allow for unrestricted access to the specimen during the epoxy debond repair. The
same instrumentation used for the original test was reconnected for use in the testing
of the repaired specimen, however a significant number of strain gages, including the
gages used to detect separation between the concrete and composite panels, were
broken during the repair.
In order to take into account any residual strains in the gages due to the testing
of the original specimen, the strain values from the end of the cyclic loading of the
original test specimen taken from the unloaded specimen immediately prior to
removing the instrumentation, were used as the initial strain values which were present
at the start of testing for the repaired specimen. By biasing the strain gages in this
manner, the most accurate representation of the actual strains present within the
system, can be maintained.
When the linear potentiometers were reconnected to the specimen, they were
rebiased immediately prior to the static loading of the repaired specimen. The use of a
new bias value for the repaired deflections facilitated comparisons of load versus
deflection data for the original and repaired test specimens however meant that the
values for the repaired specimen were relative deflections of the test specimen, rather
than absolute values.
168
4.3
4.3.1
RESULTS FOR STATIC LOADING OF THE REPAIRED SPECIMEN

Loading Protocol
The test specimen was loaded statically and cyclically to evaluate the
effectiveness of the repair on performance and integrity. The static portion of testing
used the same loading levels used for the original static testing, however the specimen
was only cycled through the minimum threshold load value of 4.5 kN (1 kip) at the
load levels of 222.4 kN (50 kips) and 578.3 kN (130 kips) instead of at every other
load level increase as done in the original static testing of the test specimen.
Additionally, the load was increased by 88 kN (20 kip) beyond the ultimate capacity
of the original test specimen to a load of 978 kN (220 kips), which was the maximum
loading capacity of the hydraulic actuators. The loading steps for the static testing of
the repair specimen are summarized in Table 3.8.
Table 4.2 Static loading protocol for testing of repaired specimen
169
4.3.2
Overall Behavior
At each load level, the target load was held briefly to allow inspection of the
test specimen in the loaded state. No additional cracking or other visual damage was
observed on the test specimen during the static loading phase for the repaired
specimen. However, multiple small popping noises originating from the composite
side deck slab were heard at the 210 kN (934 kip) load level, which was equivalent to
9.5 times the AASHTO service load. The repaired specimen was loaded up to 978 kN
(220 kip) per actuator and the ultimate capacity of the specimen was not reached. This
load was an 88 kN (20 kip) increase over the ultimate capacity of the original test
specimen of 890 kN (200 kip), which is equivalent to 9.1 times the AASHTO service
load.
The response of the repaired FRP side under static loading is compared to the
response of the original test specimen under static loading and also to the last cycle of
the cyclic phase for the original test specimen, which was the last loading experienced
by the test specimen prior to the repair. Figure 4.4 shows the overall load versus
midspan displacement envelope plots for the FRP side. In order to more easily
compare the stiffness of the deck slab before and after the repair, the load-deflection
curves for the repaired specimen and the original specimen at the end of cyclic loading
were shifted back to the origin by removing the residual deflections, as shown in
Figure 4.5.
1000
9x AASHTO900
service load
800
Load (kN)
700
Repaired
FRP side
Original
FRP side
600
500
400
300
200
Original FRP
side, end of
cyclic loading
1x AASHTO100
service load
0
0
4
5
6
Deflection (mm)
10
11
Figure 4.4 Load versus deflection envelope plots of the original and repaired test specimen for both the FRP and RC slabs
170
1000
9x AASHTO900
service load
800
Repaired
FRP side
Load (kN)
700
Original
FRP side
600
500
400
300
200
Original FRP
side, end of
cyclic loading
1x AASHTO100
service load
0
0
3
4
Deflection (mm)
171
Figure 4.5 Load versus deflection envelope plots of the original and repaired test specimen for both the FRP and RC slabs
(Residual deflections removed from data allowing all three load-deflection plots to begin at origin)
172
As observed in Figure 4.5, the repair of the specimen restores the specimens
stiffness from the significantly degraded response shown for the original specimen at
the end of cyclic loading to a structural response comparable to the original
static loading of the test specimen. For the majority of the loading range, the slope of
the load-deflection curve for the repaired FRP side was slightly lower than the values
from the original testing of the FRP side, with the repaired specimen displaying a
more linear response throughout the entire load range.
At the 98 kN (22 kip) load level, equivalent to 1x the AASHTO HS20 service
load, the vertical midspan deflection was 0.29 mm (0.011 in) for the original FRP side
slab and was 0.58 (0.022 in) for the original FRP side at the end of cyclic loading,
whereas the deflection for the repaired FRP side slab was 0.39 mm (0.016 in). At an
actuator load of 169 kN (38 kip), which was the equivalent to 1x the AASHTO
Strength I level, the center FRP side deflection for the original specimen was 0.60 mm
(0.024 in), whereas the deflections for the original specimen at the end of cyclic
loading and for the repaired specimen were 1.32 mm (0.052 in) and 0.71 mm (0.028
in) respectively. Up till the 756 kN (170 kip) load level, the repaired specimen
deflected slightly more than the original specimen, with deflection differences
between the two tests of less than 0.3 mm (0.01 in). Beyond the 756 kN (170 kip)
load level, the original specimen exhibited greater deflection values for the remainder
of the loading levels. The midspan deflection of the FRP panel side of the repaired
test specimen at the 890 kN (200 kip) load level decreased by 24% or 1.72 mm (0.068
in) to a deflection value of 5.37 mm (0.211 in) from the 7.09 mm (0.279 in) deflection
of the original specimen at the same load level.
173
4.3.3
Stiffness Change for Repaired FRP Panel Side
The damage progression of the FRP panel side slab for the original and
repaired specimen static loadings were determined through examination of the
effective stiffness and the effective stiffness ratio, which were defined previously in
Section 3.8.3 and are repeated below:
Effective stiffness =
Pf Pi
f i
4.1
where
Pf
Pi
f
f
= Minimum threshold actuator load of a load cycle

= Deflection at load Pf , at the location under consideration
= Deflection at load Pi , at the location under consideration
The effective stiffness ratio was calculated by comparing the effective stiffness
at a load level to the effective stiffness at the baseline load as shown in Equation 4.2:
Effective stiffness ratio =

4.2
The load of 98 kN (22 kip), which represented a HS20 truck load of 73 kN

(16.4 kip) with an additional AASHTO defined 33% dynamic allowance and was also
the initial load level in the static testing, was chosen as the baseline load. The
effective stiffness value determined from the loading of the original FRP side up to the
98 kN (22 kip) load level was used as the baseline value in the calculation for
determining the effective stiffness ratios for the original FRP specimen at the end of
174
cyclic loading and for the repaired FRP side. This allowed any change in stiffness
between the original and repaired test to be taken into account.
The effective stiffness ratios for the original FRP side deck slab, the original
FRP side deck slab after the completion of the cyclic loading phases and repaired FRP
side deck slab are shown in Table 4.3, Table 4.4 and Table 4.5 respectively. The
degradation trends for the slabs are most easily observed in Figure 4.6, which plots the
effective stiffness values of the slabs versus load level.
The effective stiffness ratios for the original FRP side deck slab at the end of
cyclic loading, shown in Table 4.4, of only 0.63 at the 98 kN (22kip) load level and
0.52 at the 169 kN (38 kip) load level indicate the significant damage which had
occurred to the system. Note that the cyclic loading phase was performed on the post
failure specimen to show the structural stability of the system after exposed to extreme
event loading levels therefore the level of damage observed was expected.
The
repaired specimen shows effective stiffness values far closer to original test specimen,
which indicates the repair was effective at restoring original capacity.
Throughout the entire loading range, the repaired FRP side effective stiffness
ratio gradually decreased following an approximately linear decay, with less than 5%
variation from a linear profile. Conversely, the original FRP side effective stiffness
ratio decreased by less than 10% during the first five loading cycles and then
decreased progressively greater quantities during the final five load levels.
The
repaired FRP side exhibited an effective stiffness ratio which was initially 11% lower
than the corresponding value for the original FRP side.
175
1.2
1.0
Original
Original- end of
cyclic loading
Repaired
0.8
0.6
0.4
0.2
0.0
98
222
311
400
489
578
667
Load Level (kN)
756
845
890
Figure 4.6 Stiffness ratios of the FRP side slab at different loading phases
Table 4.3 Effective stiffness ratios of original FRP panel side slab
176
Table 4.4 Effective stiffness ratios of original FRP panel side slab, end of cyclic phase
Table 4.5 Effective stiffness ratios of repaired FRP panel side slab
177
Over the next three load levels, the difference between the original and
repaired FRP side effective stiffness values gradually increased to a maximum of 19%
greater effective stiffness ratio for the original FRP side at the 400 kN (90 kip) load
level, then decreased steadily for the remainder of the loading. At the 890 kN (200
kip) load level, the effective stiffness ratio of the repaired FRP side was 19% larger
than the effective stiffness ratio for the original FRP side.
4.3.4
Deflection Profiles
levels up to the 890 kN (200 kip) load level are presented. The deflection profiles for
the repaired specimen are directly compared to the profiles from the original specimen
in order to determine the any changes in structural response between the two systems.
Unless otherwise noted, the deflection profile figures in this section all maintain the
same deflection scale, which accommodates the maximum deflection values observed
in the FRP side deck slab at the ultimate capacity of the original test specimen, in
order provide context for each deflection value in relation to the overall deformation
of the structure.
The deflection profile of the repaired specimen along the length of the deck
above the central stem presented in Figure 4.7 shows uniform deflections in the
longitudinal direction at all load levels. Similar uniform deflection profiles are also
observed above the two outer stems and these results indicate that the vertical
deformations of the stems do not significantly affect the overall structural response of
the test specimen. The maximum deflection in the deck above the stems at the 890 kN
178
(200 kip) load level was 0.87 mm (0.03 in), which indicates that the actual
corresponding deflections at the midpoints of the slabs are approximately 0.9 mm
(0.03 in) smaller than the values given from the linear potentiometers due to the
deflection contribution from stems. Note that the maximum deflections above the
stems at maximum loading were approximately 7.7 times smaller than the
corresponding midpoint deflection of the reinforced concrete side slab.
8
7
Max original FRP side

deflection (890 kN)
Deflection (mm)
+
"0"
5
4
890 kN
578 kN
400 kN
98 kN
3
2
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 4.7 Deflection profile of repaired specimen on the deck above the central stem
The deflection profiles along the centerline of the original and repaired FRP
slab are shown in Figure 4.8 and Figure 4.9. As expected, the highest deflection
values for both the original and repaired FRP side were observed for the linear
potentiometer directly below the FRP side hydraulic actuator, with progressively
smaller deflections at distances away from this location.
179
8
7
Deflection (mm)
Max original FRP

side deflection
(890 kN)
+
"0"
890 kN
5
4
3
2
823kN
578 kN
400 kN
222 kN
98 kN
756 kN
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 4.8 Deflection profile along centerline of original FRP side deck
8
7

deflection (890 kN)
Deflection (mm)
6
890 kN
823 kN
5
4
3
2
578 kN
400 kN
222 kN
98 kN
+
"0"
756 kN
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 4.9 Deflection profile along centerline of repaired FRP side deck
180
Both tests maintain nearly symmetric deflection profiles along the length of the
specimen and differ primarily in the magnitude of the center deflection values
displayed at different load levels.
original FRP slab and the repaired FRP slab, the superimposed deflection profiles
along the centerline of the decks at different load levels are shown in Figure 4.10,
Figure 4.11, and Figure 4.12. For loadings up to the 756 kN (170 kip) load level,
which is equivalent to 7.7 times the AASHTO service load, the repaired FRP side
displayed a slightly more localized deflection profile as compared to the original FRP
side, with a higher center deflection and lower deflections away from the actuator, as
shown in Figure 4.10 and Figure 4.11.
8
deflection (890 kN)
Deflection (mm)
7
6
+
"0"
5
4
3
2
Repaired FRP side

Original FRP side
1
0
-1830
-1220
-610
0
610
1220
1830
Figure 4.10 Original and repaired FRP side deflection profiles at 400 kN load level
181
8
7

deflection (890 kN)
"0"
Deflection (mm)
5
4
3
Repaired FRP
side
Original FRP
side
2
1
0
-1830
-1220
-610
610
1220
1830
8
7
Deflection (mm)
6
5
4

deflection (890 kN)
+
"0"
Original FRP
side
Repaired FRP
side
3
2
1
0
-1830
-1220
-610
610
1220
1830
182
For the load levels between 222 kN (50 kip) and 756 kN (170 kip), the
difference in center deflections between the two tests remained within the range of
0.21 mm (0.008 in) to 0.33 mm (0.013 in). This greater center deflection value for the
repaired FRP side affected the deflection profile more significantly at lower loading
levels such as the profiles shown in Figure 4.10 for the 400 kN (90 kip) load level and
was of negligible impact at higher loads as shown in Figure 4.11, which displays the
comparably shaped original and repaired FRP side deflection profiles for the 756 kN
(170 kip) load level.
At the ultimate capacity of the original test specimen, which occurred at the
890 kN (200 kip) load level (equivalent to 9.1 times the AASHTO service load), the
center deflections as well as the overall deflections at locations away from the point of
load application were significantly greater for the original FRP side as compared to the
repaired FRP side as shown in Figure 4.12.
The higher deflection values for the
original FRP side were likely due to the punching shear damage that was first
observed at the 845 kN (190 kN) load level. The repaired specimen showed no signs
of punching shear damage during its loading.
The FRP side deflection profiles along the transverse centerline of the test
specimen are presented for the original FRP side and repaired FRP side in Figure 4.13
and Figure 4.14 respectively. The highest deflection values were observed near the
actuator locations and minimal deflection values were observed above either stem.
Both tests exhibited asymmetric deflection profiles which deflected more on the side
adjacent to the central stem.
183
8
deflection (890 kN)
890 kN
845 kN
Deflection (mm)
756 kN
"0"
5
4
3
578 kN
400 kN
222 kN
98 kN
2
1
0
0
610
1220
1830
2440
Figure 4.13 Transverse deflection profile for original composite side slab
8
deflection (890 kN)
Deflection (mm)
"0"
6
5
4
3
578 kN
400 kN
222 kN
98 kN
2
1
890 kN
845 kN
756 kN
0
0
610
1220
1830
2440
Position composite side edge of deck (mm)
Figure 4.14 Transverse deflection profile for repaired composite side slab
184
The superimposed transverse deflection profiles were examined at different
loads and the profiles for the 756 kN (170 kip) load level and 890 kN (200 kip) load
level are shown in Figure 4.15 and Figure 4.16 respectively. Both tests maintained
comparable transverse deflection profiles up to the 756 kN (170 kip) load level,
equivalent to 7.7 times the AASHTO service load, with the repaired FRP side
displaying slightly more localized deflection profiles as shown in Figure 4.15. At
loads above the 756 kN (170 kip) load level, the original FRP side deflected
significantly more than the repaired FRP side as shown in Figure 4.16.
8
deflection (890 kN)
Deflection (mm)
"0"
6
Repaired FRP
side
Original FRP
side
5
4
3
2
1
0
0
610
1220
1830
2440
185
8
deflection (890 kN)
Repaired
FRP
Repaired
side
Deflection (mm)
Original FRP
side
5
"0"
4
3
2
1
0
0
610
1220
1830
2440
The envelope plots of load versus deflection for linear potentiometers adjacent
to the central linear potentiometer locations are shown in Figure 4.17 and Figure 4.18.
Figure 4.17 compares the load-deflection plots of the linear potentiometers that are
located a distance of 610 mm (2 ft) away from the center of the test specimen along
the centerline of the FRP side slab. The very similar curves for all four linear
potentiometers indicate a very symmetric deflection profile for the composite side
deck in the longitudinal direction (comparison of C3B and D3B) and also indicate that
the original and repaired FRP side deck slab exhibit very comparable responses at
distances away from the point of load application in the longitudinal direction.
186
repaired C3B
original D3B
repaired
D3B
original C3B
C3B
Load (KN)
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
D3B
0
3
4
5
Deflection (mm)
Figure 4.17 Envelope plot for linear potentiometers along the centerlines of the
original and repaired FRP slab
(Located 610 mm away in the longitudinal direction from the test specimen midspan)
The load versus deflection envelope plots for linear potentiometers which
along the centerline of the test specimen and are positioned 610 mm (2 ft) on either
side of the center of the FRP side actuator are shown in Figure 4.18. For the repaired
specimen, the two linear potentiometers remained comparable to within the tolerance
of the linear potentiometers until approximately the 578 kN (130 kip) load level was
achieved.
For the remainder of the repaired specimen loading, the linear
potentiometer located adjacent to the exterior stem, M2B, displayed an average of

13% lower deflection values, with a standard deviation of 1%, as compared to the
linear potentiometer located adjacent the central stem, M4B.
Conversely, the
deflection values for linear potentiometer M2B throughout the entire loading range of
the original experimental testing were an average of 16% smaller than the deflections
of M4B during the original test, with a standard deviation of 3%. At higher loading
187
levels, the repaired specimen deflections for both M2B and M4B were significantly
smaller than the corresponding deflections observed during the original testing,
indicating that the implemented repair was effective.
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
repaired M4B
original M4B
original M2B
Load (KN)
repaired M2B
M2B M4B
0
3
4
5
Deflection (mm)
Figure 4.18 Envelope plot for linear potentiometers on original and repaired FRP side
along the centerline of the test specimen
(Located 610 mm away in the transverse direction from the FRP side centerline)
The deflections of the original and repaired FRP slabs were compared both
near and away from the point of load application and in terms of longitudinal and
transverse deflection profiles. The maximum deflection values measured in the both
the original and repaired FRP deck slabs were located at the midspan of the slabs,
directly below the FRP side actuator. For the majority of the loading range, the center
deflection values for the repaired FRP side slab were slightly higher than the original
FRP slab center deflections and the deflections away from the point of load
application in both the longitudinal and transverse direction were comparable for both
tests. At higher loads, the repaired FRP slab deflected less both in terms of center
188
deflections as well as deflections at points away from the point of load application. At
the 890 kN (200 kip) load level, equivalent to over nine times AASHTO service load,
the original FRP side displayed a 7.09 mm (0.279 in) center deflection, while the
repaired FRP side displayed a 5.42 mm (0.213 in) center deflection. A detailed
discussion of the progression of the center deflections for both the original and
repaired FRP slabs, as they pertain to the effective stiffness of the slabs, is provided in
Section 3.8.3, beginning on page 87. The stable linear deflection response displayed
by the repaired FRP slab is an indication that the repair of the FRP side deck slab was
successful.
4.3.5
Strain Development
examination of strain gages applied to the steel and composite reinforcement within
the test specimen. The precise location of all strain gages is given in Section 3.6.2.
Note that strain values reported as negative indicate compressive strains whereas
positive values indicate tensile strains. Since the structural performance of the FRP
panel side slab is the focus for testing of the repaired specimen, only strains from the
FRP panel side slab and non from the RC side are shown. Performance comparisons
are made between the repaired specimen strain values and the strains from the original
testing of the FRP deck slab.
As discussed previously in Section 0, the repaired specimen strain gages were
biased using strain values from the end of the cyclic loading of the original test
specimen. This allowed for any residual strains in the gages due to the testing of the
189
original specimen to be taken into account by presenting absolute strain values rather
than relative strain values, which gives a more accurate picture of the actual strains
present within the structure. For the discussion of strain progressions, the changes in
the strain values for different load levels were presented for the repaired specimen in
terms of relative strains rather than absolute strains, which were biased from the start
of the repaired specimen testing, in order to allow for easier comparisons with the data
from the original test specimen at different load levels.
4.3.5.1 Strain Progression for Gages at Center of Composite Side Deck

At the 1x AASHTO service load of 98 kN (22 kip), all composite side deck
relative strains in the original test specimen were less than 100 microstrains and were
less than 130 microstrains for the repaired specimen, with the highest strains measured
in the strain gages directly below the two actuators. At the 400 kN (90 kip) load level,
which is equivalent to approximately four times the AASHTO service load, the
highest FRP deck side strain value of 516 microstrains was recorded for the bottom
transverse FRP side strain gage at midspan from original test specimen whereas the
equivalent relative strain value for the repaired test specimen was 577 microstrains.
The bottom longitudinal relative strain gage readings on the strain gage located
directly below the composite actuator, was 241 for the original test specimen and 531
for the repaired test specimen for this load level.
At ultimate capacity for the original test specimen, which occurred at load of
890 kN (200 kip), equivalent to approximately nine times the AASHTO service load,
190
the highest overall strains in the FRP panel side deck slab of the original test specimen
and repaired test specimen were observed in the bottom strain gages directly below the
composite actuator. The bottom transverse and longitudinal strains measured on the
strain gages attached to the bottom of the FRP panel, directly below the composite
side actuator were 920 microstrains and 2442 microstrains respectively for the original
test specimen and were 1193 microstrains and 1566 microstrains respectively for the
repaired test specimen. The strains measured in the bottom transverse rebar located
along the test specimen centerline and positioned on the composite side a distance of
152 mm (6 in) away from the edge of the center stem displayed were -1676
microstrains for the original test specimen and 1728 microstrains for the repaired test
specimen.
Load versus strain envelope plots, which include both the original and repaired
test specimen strain values, are presented for the FRP side strain gages located directly
below the actuator. Tabulated values at each load level are also shown for these strain
gages.
The values for the repaired specimen are reported in the tables both as
calculated using the biasing method described in Section 0, which takes into account
the residual strains present in the specimen due to the original testing, and as
calculated using the start of the repaired specimen testing as the bias value, which does
not take into account any previous deformations and thus provides relative strain
values.
The first method of biasing the repaired specimen strains gives a more
accurate picture of the actual strains within the specimen, whereas the second biasing
method provides the ability for more direct comparisons between the original and
191
repaired test specimen of relative changes in strain values over the course of the static
loadings.
The load versus strain envelope plot of the bottom longitudinal strain gage
located directly below the composite side actuator (CM3L) is shown in Figure 4.19 for
the original and repaired test specimen as well as the original specimen at the end of
the cyclic loading phase. As observed in Figure 4.19 and Table 4.6, which shows the
strain values at each load level, the initial slope of the load versus strain plot for the
original test specimen was approximately twice as steep as the slope for the repaired
specimen original specimen.
1000
9x AASHTO 900
service load 800
Load (KN)
700
Original
CM3L
600
Repaired
CM3L
CM3L
500
400
300
200
Residual
Strain
1x AASHTO 100
service load
0
0
500
Original CM3L, end

of cyclic loading
1000
1500
2000
2500
Figure 4.19 Original and repaired test specimen load versus strain plots for the bottom
longitudinal FRP side strain gage, CM3L
192
Table 4.6 Strain values for CM3L at different load levels
This relationship was maintained up until a load of approximately 400 kN (90

kip), which is over four times the AASHTO service load, after which the original
specimen displayed a significantly more relaxed slope.
The repaired specimen
maintained a nearly linear response for the CM3L gage, especially for loadings
beyond the 222 kN (50 kip) load level. While the repaired specimen increased in
strain at a significantly slower rate for all loads above 400 kN (90 kip), the repaired
specimen maintained higher strain values until a load of 762 kN (171 kip) was reached
due to the residual strain of 543 microstrains present within the strain gage from
previous testing. The total change in strain values from initial loading to ultimate
capacity for the repaired specimen is 36% less than equivalent value for the original
specimen at 1565.8 microstrains for the repaired specimen versus 2441.2 microstrains
for the original specimen.
193
The load versus strain envelope plots for the original and repaired specimens
bottom transverse strain gage attached to the center of the FRP panel are shown in
Figure 4.20 and tabulated in Table 4.7 which shows the strain values at each load
level. The slope of the curves for both tests remain similar up to the 400 kN (90 kip)
load level, which is approximately four times AASHTO service load, with less than
6% difference between the strain values for the two tests at this load level. For higher
loadings, the original specimen exhibits a more nonlinear response, which is likely
related to the onset of the separation between the FRP panels and the concrete that was
detected at approximately the 489 kN (110 kip) load level, equivalent to 5.0 times the
AASHTO service load. The more linear response observed during the loading of the
repaired specimen is an indicator of that the FRP panel repair was successful in
improving the structural response of the FRP side deck slab.
1000
9x AASHTO 900
service load 800
Original
CM3T
Load (KN)
700
Repaired
CM3T
CM3T
600
500
400
Original CM3T,
end of cyclic
loading
300
200
1x AASHTO 100
service load
0
Residual Strain
0
500
1000
1500
2000
2500
Figure 4.20 Original and repaired test specimen load versus strain plots for the bottom
transverse FRP side strain gage, CM3T
194
Table 4.7 Strain values for CM3T at different load levels
The loads versus strain envelope plots for the original and repaired test
specimens top longitudinal strain gage are shown in Figure 4.21 and the strain values
at each load level are shown in Table 4.8. The positive sign of the data on gage
DTL8-3 for both the original and repaired specimen is indicative of more localized
deformations in the region directly below the FRP side actuator, where strain gage
DTL8-3 is located. A residual strain of 276 microstrains was present on the repaired
strain gage due to the previous tests. As stated previously, the more linear response
observed during the loading for the repaired specimen as compared to the original
specimen is a positive indicator for the performance of the FRP panel repair.
The load versus strain responses for the two tests top transverse strain gage
located directly below the FRP panel side actuator (Figure 4.22) illustrates a
significantly different strain response by the DTT10-3 strain gage for the repaired
195
specimen as compared to the original specimen. While the strain in DTT10-3 for the
original specimen increased in compression with increasing load levels and exhibited a
discontinuity in strain response at the 845 kN (200 kip) load level, the repaired
specimen increases slightly in compression values up to approximately the 222 kN (50
kip) load level, then gains positive, or tensile, strain values for the remainder of the
loading. Since the top longitudinal strain gage, DTL8-3, displayed tensile values
thereby indicating the presence of localized deformations, it is a reasonable
extrapolation that the transverse direction strain response changes due to an increase in
the level of localized deformations in the region directly below the actuator.
1000
9x AASHTO 900
service load 800
Load (KN)
700
DTL8-3
600
Original
DTL8-3
500
400
Repaired
DTL8-3
300
200
1x AASHTO 100
service load
0
-1500
Original DTT10-3, end of

cyclic loading
-1000
-500
0
Residual
Strain
500
1000
Figure 4.21 Original and repaired test specimen load versus strain plots for the top
longitudinal FRP side strain gage, DTL8-3
196
Table 4.8 Strain values for DTL8-3 at different load levels
1000
9x AASHTO 900
service load 800
Load (KN)
700
DTT10-3
Repaired
DTT10-3
600
500
400
300
Original
DTT10-3
200
Original DTT10-3,
Residual
1x AASHTO 100
end of cyclic loading
Strain
service load
0
-1500
-1000
-500
0
500
1000
Figure 4.22 Original and repaired test specimen load versus strain plots for the top
transverse FRP side strain gage, DTT10-3
197
Table 4.9 Strain values for DTT10-3 at different load levels
4.3.5.2 Strain Profiles along the Longitudinal Direction of the Deck

It is important to determine the degree to which the strains due to live loading
distribute along the deck in both directions. The strain profiles along the longitudinal
directions for the centerlines of the composite side deck slab are shown for the original
and repaired tests in Figure 4.23 and Figure 4.24 respectively. Both sides maintained
similarly shaped strain profiles with the point of inflection, where the strain values
change from tension to compression strains, positioned in approximately the same
location. The primary difference between the two tests stems from the degree of
spread in the strain values for gage DTL8-3, which is located directly below the FRP
side actuator.
198
3000
2500
890 kN
2000
845kN
+
"0"
756 kN
1500
578 kN
1000
500
400 kN
222 kN
98 kN
-500
-1000
-1830
-1220
-610
610
1220
1830
Figure 4.23 Original specimen strain profile for gages attached to bottom of FRP
panel along line 3
(Located directly along the FRP side centerline)
2500
890 kN
845 kN
756 kN
578 kN
400 kN
2000
1500
1000
+
"0"
222 kN
98 kN
0 kN
500
0
-500
-1000
-1830
-1220
-610
0
610
1220
1830
Figure 4.24 Repaired specimen strain profile for gages attached to bottom of FRP
panel along line 3
(Located directly along the FRP side centerline)
199
The repaired specimen started out with approximately 550 residual strains,
however it exhibited an overall change in strain of over 35% less than the original
specimen values over the course of the static loading. The progression of these central
longitudinal strain gages were discussed in detail in Section 4.3.5.1, which can be
referenced for further information.
The strain profiles for gages attached to the top longitudinal steel
reinforcement along the centerline of the composite are shown in Figure 4.25 and
Figure 4.26. The plot for the original test shows a similar profile shape to the bottom
longitudinal profiles for the FRP side deck slab, with tensile strains at the point of load
application and a point of inflection at a position prior to the strain gage located 610
mm (2 ft) away from the centerline of the test specimen. The repaired specimen has a
modified shape, with the outermost strain gage reading tensile values of approximately
150 to 200 microstrains rather than nearly zero strain values as shown for the original
test specimen.
This change in profile shape is due the residual strain of 206
microstrains which was present on the strain gage from the start of the repaired
specimen testing. Note, however, that the top longitudinal strain profile maintains the
same signs as the bottom longitudinal strain profile such that the center strain gage is
in tension and the two farther away strain gages are in compression, which is
indicative of more localized deformations in the region. This phenomenon will be
addressed further in the analysis sections.
200
3000
2500
890 kN
2000
845 kN
+
"0"
756 kN
1500
578 kN
1000
400 kN
500
222 kN
98 kN
-500
-1000
-1830
-1220
-610
610
1220
1830
Figure 4.25 Strain profile for original specimens top longitudinal rebar along line 8
3000
2500
+
"0"
2000
1500
1000
890 kN
578 kN
400 kN
98 kN
500
0
-500
-1000
-1830
-1220
-610
0
610
1220
1830
Figure 4.26 Strain profile for repaired specimens top longitudinal rebar along line 8
201
4.3.6
The occurrence and extent of any separation between the FRP panels and the
concrete which was poured on top of the panels was evaluated through the
examination deflection data from instrumentation on the FRP side deck slab.
Unfortunately, the strain gages necessary for determining the separation between the
FRP panel and concrete from strain data were broken during the implementation of the
repair and so the discussion of panel separation is limited to deflection information.
Differences in deflection values obtained from pairs of linear potentiometers located
above and below the FRP panel side deck slab, were used to obtain information
regarding the level of separation, if any that had occurred. A differential between the
top and bottom pot readings indicated that the concrete and composite were no longer
acting in unison and were now deflecting independent of each other.
The load versus deflection envelope plots of the top and bottom pair of linear
potentiometers located along the transverse centerline of the test specimen on either
side of the actuator on the FRP panel deck are shown in Figure 4.27 and Figure 4.28.
Based on the observed fluctuations in measured deflections for top and bottom linear
potentiometer pairs on the reinforced concrete side deck during the testing of the
original specimen, a tolerance level of 0.16 mm (0.006 in) was established as the
threshold value for determining that a difference between the top and bottom
deflections was significant.
For the pair of linear potentiometers located near the outer stem shown in
Figure 4.27, a difference of 0.17 mm (0.007 in) between upper and lower deflection
values was observed at a load of 835 kN (188 kip), which is approximately 8.5 times
202
the AASHTO service load. At the 890 kN (200 kip) load level, which was the
ultimate capacity of the original test specimen, the difference between the linear
potentiometers shown in Figure 4.27 increased to 0.22 mm (0.009 in) and at 980 kN
(220 kip), the maximum capacity of the actuators, the difference increased to 0.42 mm
(0.017 in).
This separation was significantly less than the corresponding values
measured during the original test, which had a separation at ultimate capacity of 1.09
mm (0.043 in).
In addition, the first detectable separation was recorded for the
original test at a load of 673 kN (151 kip), which was 162 kN (36 kip) lower than the
corresponding load for the repaired specimen.
1000
900
800
700
600
500
400
300
200
100
0
Bottom
Load (KN)
Top
3
4
5
Deflection (mm)
Figure 4.27 Load versus deflection plots for linear potentiometers on top and bottom
of repaired FRP side deck
(Located along test specimen centerline and 610 mm away from the composite side
centerline, 178 mm from the inner edge of the outer stem)
203
1000
900
800
700
600
500
400
300
200
100
0
Bottom
Load (KN)
Top
3
4
5
Deflection (mm)
of repaired FRP side deck
(Located along test specimen centerline and 610 mm away from the composite side
centerline, 152 mm from the FRP side edge of the center stem)
The upper and lower deflection values for the pair of linear potentiometers
located near the center stem shown in Figure 4.28 were separated by 0.17 mm (0.007
in) at a load of 356 kN (80 kip), which is equivalent to 3.6 times the AASHTO service
load.
The difference between the upper and lower linear potentiometers for the
repaired FRP test increased to 1.13 mm (0.044 in) at 890 kN (200 kip) and 1.47 mm
(0.058 in) at the ultimate capacity of the actuators. For the original test specimen, the
corresponding upper and lower linear potentiometers were separated by 0.17 mm
(0.007 in) at a load of 489 kN (109 kip) and the difference increased to 3.12 mm
(0.123 in) at ultimate capacity. While the onset of detectable separation for the
repaired specimen occurred at a load which was 133 kN (30 kip) lower than that for
the original specimen, the separation for the repaired specimen remained over 2.75
204
times smaller than for the original specimen. This indicates that the epoxy repair was
effective at reinstating the bond between the FRP panels and the adjacent concrete and
perhaps creates a better bond that under original conditions. This also indicates that
surface treatment during original construction to create a better bond between the
concrete and the FRP panels may be desirable.
4.4
4.4.1
RESULTS FOR CYCLIC LOADING OF THE REPAIRED SPECIMEN

Loading Protocol
The cyclic testing of the repaired specimen consisted of 500 cycles at 168 kN
(37.8 kip), which is equivalent to 2x the AASHTO service load of 73 kN (16.4 kip)
with the additional 15% fatigue with an additional AASHTO defined 15% dynamic
allowance for the fatigue limit state [91]. After the cyclic loading was completed, the
specimen was loaded with a single cycle static loading to 978 kN (220 kips) in order
to compare system structural response from before and after the cyclic loading.
Table 4.10 Cyclic loading protocol for repaired specimen
205
The current cyclic loading protocol was different from the cyclic loading of the
original test specimen, which consisted of 1,000 cycles at 1x AASHTO service load
followed by 1,000 cycles at 2x AASHTO service load. The cyclic loading of the
original specimen at 1x service load caused no noticeable change in structural
response to the system and over 80% of the changes in stiffness for the FRP panel side
deck occurred in the first 500 cycles of the 2x service load phase. Therefore, it is
reasonable to directly compare the FRP panel side deck performance during the 2x
AASHTO service cyclic loading from the repaired specimen with the first 500 cycles
of the 2x AASHTO service loading of the original test specimen.
In order to determine the overall response of the structure, the data from the
cyclic testing was biased using the original bias scan from the start of static testing of
the repaired specimen, enabling the cumulative deflections throughout the loading of
the repaired test specimen to be established. Therefore, a comparison of stiffness
values would be more accurate than a comparison of the absolute magnitude of
deflection values.
4.4.2
Changes in Center Deflections during Cyclic Loading
The effective stiffness ratios for the original FRP panel side deck during the
first half of cyclic phase II and the repaired FRP panel side deck during the cyclic
loading were calculated using the adaptive secant stiffness method described on page
87 in Section 3.8.3 and are shown in Table 3.21 and Table 3.22 respectively. Since
the load acting on the system remains constant throughout the cyclic phase, any
change in the stiffness ratio is attributed to a change in deflections. The progression of
206
the effective stiffness ratios for both sides over the course of 500 cycles of 2x
AASHTO service load is shown in Figure 3.79.
The effective stiffness ratios for the original FRP panel side, shown in Table 3.21,
dropped by 12% during the first 250 cycles and displayed a total decrease in effective
stiffness of 14% during the first 500 cycles of cyclic phase II for the original test
specimen. Conversely, the cyclic phase for the repaired specimen had no measurable
impact on the specimen, which maintained its full effective stiffness throughout the
loading, with less than 2% variation throughout the loading phase.
1.00
0.80
Repaired FRP
panel side
0.60
Original FRP
panel side
0.40
0.20
0.00
0
125
250
Cycle #
375
500
Figure 4.29 Effective stiffness versus cycle # for original and repaired FRP side deck
207
Table 4.11 Effective stiffness ratios for FRP panel side of original test specimen
during the 1st 500 cycles of cyclic loading phase II (2x AASHTO service load)
Table 4.12 Effective stiffness ratios for FRP panel side of repaired test specimen
during cyclic loading (2x AASHTO service load)
208
4.4.3
Changes in Strains during Cyclic Loading
The progression of strains during the cyclic loading phases for important strain
gages in the deck slab is presented in Table 3.23. Note that all gages were biased
using the original strain gage values from the start of the static testing in order to
record the complete history of the strain imposed on each gage due to all previous
loading. Also, all data shown in the tables below were taken with the specimen at the
full target load for a given cyclic phase.
Throughout the cyclic loading of the repaired specimen, only minor progressions in
strain values were observed for the FRP side deck strain gages. For the strain gages
located directly below the FRP side actuator which are shown in Table 3.23, the strain
values for the repaired specimen changed by less than 15 microstrains or 4% over the
course of the cyclic phase and changed by less than 20 microstrains or 5% for the
original specimen over the course of the first 500 cycles of cyclic phase II. The
consistent strain values observed for the FRP panel side during the cyclic loading
phases, both for the original specimen and for the repaired specimen, are an indication
that the FRP panel system exhibits a very stable response under cyclic loading.
Table 4.13 Strain progresssion for composite side gages under target loadslocated directly below the composite side actuator
209
210
4.4.4
Changes in Separation between the FRP Panel and the Concrete
The progression of the separation between the concrete and the composite at
the same locations examined in Section 4.3.6 is summarized in Table 3.26. During the
first 500 cycles of cyclic phase II for the original specimen, the pair of linear
potentiometers located near the central stem continued to display very similar
separation values whereas the linear potentiometer pair near the outer stem exhibited a
minor change of 0.19 mm (0.008 in) in the overall separation distance between the
concrete and the FRP panel. Both pairs of linear potentiometers for the repaired
specimen show no change in separation values over the course of the cyclic loading
phase.
Table 4.14 Separation progression for linear potentiometers located along the test speciment centerline, near the stems
(See Section 5.2.2 for detailed information on precise location of each linear potentiometer)
211
212
4.4.5
Static Loading Following the Repaired Specimen Cyclic Phase
Apart from a minor increase in residual deflection, the structural response of

the repaired FRP panel side deck was equivalent before and after the cyclic loading
phase, as shown in Figure 4.30. This indicates that minimal damage was accrued due
to the cyclic loading phase of the repaired FRP side deck.
1000
9x AASHTO 900
service load 800
700
600
500
400
300
200
1x AASHTO 100
service load
0
Load (KN)
Repaired FRP
side
3
4
5
Deflection (mm)
Repaired
FRP
Original FRP
side after
cyclic
side
loading
Figure 4.30 Envelope plot of load versus deflections of repaired FRP side before and
after cyclic loading phase
4.5
SUMMARY
After the deck was loaded to ultimate capacity, a field applicable epoxy
injection repair technique was implemented to reinstate load transfer between the FRP
panels and the concrete, which had displayed initial signs of separation during the
quasi static loading of the test specimen at a load of approximately five times the
213
AASHTO service. The repaired specimen was retested both statically and cyclically
and its structural response was shown to be very comparable to the structural response
of the original test specimen.
Since only the FRP panel side was repaired,
performance comparisons between the original testing of the FRP structural formwork
system and the repaired FRP structural formwork system were addressed. Aspects of
the response of the FRP panel reinforced system including the overall response, the
response between the panels, changes in stiffness, and the progression of damage
within the system were evaluated and discussed.
The repaired specimen displayed slightly larger center deflection values at
lower load levels, however exhibited less change in its effective stiffness as compared
to the original test specimen. The overall load carrying capacity of the specimen was
increased beyond the original specimens 890 kN (200 kip) ultimate capacity and the
system still had reserve strength at the maximum applied load of 978 kN (220 kip).
Initial signs of separation between the concrete and the FRP panels were observed at
load levels similar to the original test specimen, however significantly smaller
separation values were recorded for the repaired specimen and no deterioration of
performance was observed due to the onset of initial separation.
The repaired
specimen showed no change in performance during or after the application of cyclic

loading. The very linear response of the repaired specimen throughout the loading
levels and the stable response during the cyclic phase indicated that the implemented
repair was performed as desired.
Since the composite panel was not damaged during loading, an efficient repair
system could be utilized. The efficacy of the system is that that the composite fails at
214
a much higher load than the damage of the concrete, thereby allowing for a relatively
inexpensive repair of the concrete to take place rather than requiring a more costly
replacement of damaged reinforcement. The overdesign of FRP panel system allows
for robust design that exhibits a stable even after ultimate capacity is reached and
displays a readily observable failure mode.
CYCLIC RESPONSE OF THE FRP PANEL STAY-IN-PLACE

STRUCTURAL FORMWORK SYSTEM
5.1
INTRODUCTION
This chapter discusses the testing of a second box girder bridge specimen with
the hybrid fiber reinforced polymer (FRP)-concrete bridge deck system. The testing
of this test specimen focuses on the behavior of the system under repeated cyclic
loading. A test specimen with the same external geometry and setup of prefabricated
composite panels that act as both the formwork and the lower level of reinforcement
for the bridge deck as the two cell box girder specimen described in Chapters 3 and 4
is utilized. In the same manner as the first test specimen, a direct performance
comparison is drawn between the FRP panel strengthened concrete system and a
traditional steel reinforced concrete system.
5.2
OVERALL GEOMETRY AND CONSTRUCTION DETAILS
The overall geometry of the second specimen was the same as the first test
specimen, which consisted of a two-cell box girder bridge segment with a center-tocenter distance of 1830 mm (6 ft) between each of the stems and a longitudinal length
of 3660 mm (12 ft) as shown in Figure 3.2. For further details on overall specimen
design and construction, refer to pages 43-52 in Section 3.2. The differences between
the first and second test specimens, as summarized in Table 5.1, relate to the detailing
used to connect the FRP panels to the surrounding bridge system and the manner in
215
216
which the specimen is tested. The loading protocols for the static and cyclic phases
are discussed in Sections 5.6.1 and 5.7.1 respectively.
The bottom slab and lower portion of the stems for the first and second test
specimens were poured at the same time, therefore the properties for the steel
reinforcement and 1st concrete pour can be found in Table 3.1 and Table 3.2
respectively. The concrete for the deck slab and upper portion of the stems had a 28day compressive strength of 33.5 MPa (4.86 ksi) and a compressive strength on the
day of test of 33.9MPa (4.92 ksi). The strength development of the deck concrete is
shown in Figure 5.1 below.
Table 5.1 Comparison of 1st and 2nd test specimens
217
Concrete Strength (MPa)
40
35
30
25
Day of
test
20
15
10
5
0
0
14
21
Days
28
35
42
Figure 5.1 Strength development for deck slab concrete in test specimen #2
Figure 5.2 Construction of test specimen formwork for 2nd pour
218
5.3
PANEL PLACEMENT AND CONNECTION DETAILING
Using the same configuration as Test Specimen #1, three FRP panels were
used side by side in the composite side deck of Test Specimen #2 to check the
effectiveness of load transfer between the panels.
The central FRP panel with
dimensions of 1220 mm (4 ft) wide, 1830 mm (6 ft) long, and 6.3 mm (0.25 in) thick
was positioned in the middle of the FRP strengthened side of the specimen and was
bordered on both sides by narrower panels, each 610 mm (2 ft) wide as shown in
Figure 5.3.
The connection detailing used in this second test more closely modeled field
repair construction methods and aimed to better accommodate protruding stem rebar.
The presence of bent stem rebar is encountered in repair of preexisting bridges where
only the deck surface is replaced and existing stem rebar must be integrated into the
new deck to enable proper load transfer. The location of the protruding stem rebar
was measured and grooves with dimensions 139.7 mm (5.5 in) long by 25.4 mm (1 in)
wide were cut into the panel to accommodate the stem rebar. These grooves were
located at an average spacing of 305 mm (12 in) on center, which corresponds to the
spacing of the vertical stem rebar.
The ends of each groove were rounded by
terminating each groove cut with a drilled hole in order to reduce stress concentrations
in the FRP panel. End connection details for the first and second test specimen are
shown in Figure 5.4, parts a) and b) respectively.
219
Figure 5.3 Schematic of FRP panel locations (same overall locations for both 1st and 2nd test specimens)
220
(a) End condition detail from Test Specimen #1.
(b) End condition detail from Test Specimen #2

Figure 5.4 End connection detailing of FRP panels
221
Additional holes were drilled in the end regions of the panels so that 12.7 mm
(0.5 in) diameter threaded rods could be used to secure the panels to the girders prior
to the pouring of the deck concrete. The holes in the FRP panels were used as
templates for ensuring proper alignment of the corresponding holes drilled into the
stem concrete as shown in Figure 5.5. After all holes were drilled into the concrete,
any loose concrete pieces were vacuumed up and each threaded rod was attached to
the stem concrete using a twocomponent 1:1 ratio epoxy-based adhesive called
Simpson Epoxy-Tie Set 22. This adhesive system is designed for use as a high
strength, non-shrink anchor grouting material and its properties are summarized in
Table 5.2 [157]. The threaded rods were held to the panels by wrench tightened nuts
and washers installed on both sides of the panels as shown in part (b) of Figure 5.4.
Figure 5.5 Drilling of stem concrete using FRP panels for alignment of holes
222
Table 5.2 Properties of Simpson Epoxy-Tie adhesive system [157]
Placement of the threaded rod anchors in relation to the stem concrete was
determined by maximizing the load capacity of the anchors with respect to the
breakout strength of the concrete in shear, the shear pullout strength of the FRP panel
and the tensile pullout of the threaded rod from the stem concrete.
The necessary embedment depth to ensure that the threaded rods would yield
prior to a tensile pullout from the stem concrete due to failure of the epoxy was
calculated as:
Lembed =
F
( d ) adhesive
5.1
where
Lembed = Minimum embedment depth required to yield rod prior to epoxy failure
d
= Force required to yield rod = y
2
= Yield strength of rod
= Diameter of rod
223
adhesive = Bond strength of adhesive

The value for the minimum required embedment length calculated from
Equation 5.1 was 57 mm (2.23 in), which was smaller than the manufacturer
recommended value of 108 mm (4.25 in) [157], therefore the manufacturer
recommended embedment depth was used. The position of the threaded rod anchor in
relation to the FRP panel and the concrete stem is illustrated in Figure 5.6 below.
As shown in Figure 5.6, the threaded rods were placed such that a clear
distance of 60 mm (2.38 in) from the edge of the FRP panel, labeled as distance L,
and a clear distance of 76 mm (3 in) between the threaded rod and the edge of the stem
concrete, labeled as distance ca1, was achieved. The distances L and ca1 were
determined by calculating the optimal distance for the anchor to be placed away from
the edge of the FRP panel in order to maximize the concrete breakout strength in shear
for a single anchor in cracked concrete and the maximum shear force that can be
applied to the FRP panel prior to inducing a pullout failure in the FRP.
le
FRP Panel
c a1
Anchor
Concrete Stem
Figure 5.6 Detail of anchor position with respect to concrete stem and FRP panel
224
The concrete breakout strength in shear of a single anchor in cracked concrete

was calculated using Equation D-24 from the ACI SCM-31 [158]:
l
Vb = 7 e
do
0.2
do
f c' (c a1 )
1.5
5.2
where
Vb = Concrete breakout strength (psi)
le
= Load bearing length of anchor in shear (in)
d o = Diameter of anchor (in)
f c'
ca1
= Concrete strength (psi)

= Distance to edge of concrete (in)
The maximum shear force that can be applied prior to pullout failure in the
FRP panel was calculated as:
V = A 12
5.3
where
V
= Concrete breakout strength
A = Area in shear, A = 2tL
t
= Thickness of composite
L
= Distance from hole to edge of the FRP panel
12 = Shear strength of composite
After the FRP panels were lowered into place, as shown in Figure 5.7, the steel
reinforcement in the deck was positioned and tied. Figure 5.8 shows the as built deck
just prior to the concrete pour. Note that wooden shored formwork was installed for
the entire deck except for the region in which the FRP panels are placed.
225
Figure 5.7 Composite side deck after FRP panel installation
Figure 5.8 Reinforcement system in deck, as built
226
5.4
SPECIMEN SETUP AND INSTRUMENTATION
The current test specimen setup was the same as the test setup for the first test
specimen, with the specimen supported on six load cells and tested using two actuators
spaced 1830 mm (6ft) apart from each other and in the center of the two deck slabs.
Figure 5.9 shows the as-built test specimen with instrumentation, prior to the
installation of the actuators. Information regarding the specimen setup can be found in
Section 3.5.
Figure 5.9 Test specimen with instrumentation, prior to installation of actuators
The instrumentation used on this test specimen is similar to what was used on
the first specimen. The position of some strain gages and linear potentiometers were
moved in order to better characterize the structural response of the system. Table 3.6
summarizes the total quantity of instrumentation used for this test specimen and the
approximate locations in which each instrumentation type was used.
227
Table 5.3 Summary of instrumentation used for test specimen #1
5.4.1
Linear Potentiometers and Inclinometers
Linear potentiometers were used to measure the deflections of the test

specimen and determine the shape of the structure under the applied load. The
locations of the linear potentiometers placed on the underside and top side of the deck
are shown in Figure 5.10 and Figure 5.11 respectively. Selected pairs of linear
potentiometers were placed in pairs with one linear potentiometer on the top side of
the deck and the other located on the underside of the deck in order to evaluate any
separation between the concrete and the FRP panels that may occur during loading.
Horizontal linear potentiometers and the inclinometers were used to measure
changes in slope within the deck and determine more information regarding the end
conditions of the deck slab. The layout of the horizontal linear potentiometers and the
inclinometers, which was the same as for test specimen #1, is shown in Figure 3.22.
228
Figure 5.10 Layout of linear potentiometers located on the underside of the composite deck
229
Figure 5.11 Layout of linear potentiometers located on the top of the composite deck
230
5.4.2
Strain Gages
A total of 121 strain gages were attached to the steel reinforcement throughout
the test specimen as well as to the top and bottom side of the FRP panel. The precise
layouts of the strain gages are shown in Figure 5.12 through Figure 5.19. A brief
summary of the distribution of strain gages throughout the specimen is provided in
Table 5.4.
Table 5.4 Strain gages used in test specimen #2
The positions of the strain gages attached to the bottom side of the FRP panel
side deck are shown in Figure 5.12 and Figure 5.13 , oriented in the longitudinal and
transverse directions, respectively. The strain gages attached the bottom of the FRP
231
deck were spaced 305 mm (1 ft) apart from each other in the transverse direction and
approximately 610 mm (2 ft) apart from each other in the longitudinal direction.
Figure 5.12 Layout of longitudinal strain gages attached to bottom of FRP side deck
The longitudinally oriented strain gages shown in Figure 5.12 were positioned
in two rows along the FRP deck slab, with one longitudinal row at the midspan of the
slab and the other offset from the slab centerline by 305 mm (12 in) in the direction of
the outer stem. These longitudinal lines of strain gages serve to assess the progression
of strains at different distances away from the applied load. Strain gages were located
directly on either side of the intersection between FRP panels in order to evaluate the
effectiveness of stress transfer between adjacent panels.
The transversely oriented
strain gages attached to the bottom of the FRP side deck consisted of two long rows of
232
gages as shown in Figure 5.13, with one row along the test specimen centerline and
another row located 610 mm (2 ft) away from the specimen centerline.
Figure 5.13 Layout of transverse strain gages attached to bottom of FRP side deck
In order to further investigate the structural response of the FRP panels, strain
gages were attached in different locations throughout the top side of the central FRP
panel. The two longitudinal strain gages shown in Figure 5.14 were positioned along
the midspan of the FRP side deck at the same location as gages that were applied to
the bottom of the FRP panel, therefore allowing for direct comparison of strain values
between the top and bottom side of the FRP panel. The transversely oriented strain
233
gages shown in Figure 5.15 were attached along the midspan of the FRP deck, both in
the middle of the flat panel sections as well as on top of the FRP panel stiffeners, and
near the edges of the FRP panel.
Figure 5.14 Layout of longitudinal strain gages attached to top of central FRP panel
Figure 5.15 Layout of transverse strain gages attached to top of central FRP panel
234
Strain gages were also attached to the top and bottom layer of steel
reinforcement in the deck slab and were attached to both longitudinally and
transversely oriented rebar. Two strain gages were attached to the top longitudinal
rebar in a row that was as close as possible to the centerline of the two deck slabs.
The middle strain gage in each row was placed at the centerline of the test specimen
and the second strain gage was spaced a distance of 610 mm (2 ft) increments along
the bar as shown in Figure 5.16. The top transverse steel rebar that ran along the
specimen centerline was instrumented with nine strain gages as shown in Figure 5.17.
Figure 5.16 Layout of strain gages attached to top longitudinal steel reinforcement
(same layout of strain gages attached to the bottom longitudinal reinforcement)
235
Figure 5.17 Layout of strain gages attached to top transverse steel reinforcement
The 27 strain gages attached to the bottom layer of transverse rebar were
primarily located on three lines of strain gages, which each contained eight strain
gages as shown in Figure 5.18. Three additional bars with one strain gage each added
further instrumentation to the reinforced concrete side. The bottom longitudinal rebar
in the deck were instrumented with a total of two strain gages, located in the same
236
position as the strain gages on the top longitudinal steel reinforcement, as shown in
Figure 5.16.
Figure 5.18 Layout of strain gages attached to bottom transverse steel reinforcement
The strain gages attached to the vertical steel reinforcement in the stems are
arranged in one of two patterns, termed A and B, and are located at different
positions within the test specimen cross section as shown in Figure 5.19. Strain gage
pattern A contains eight gages, with six attached to the outer stem of the FRP side and
237
two attached to the center stem as shown in Figure 5.20. Strain gage pattern B
contains three strain gages total, with one strain gage attached to the outer FRP side
stem and two gages attached the center stem as shown in Figure 5.21. There are a
total of 48 strain gages attached to the vertical stem rebar.
Figure 5.19 Strain gage pattern locations for gages attached to vertical stem rebar
238
Figure 5.20 Strain gage pattern A

(Consists of six strain gages attached to the vertical stem rebar of the FRP side outer
Figure 5.21 Strain gage pattern B

(Consists of one strain gage attached to the vertical stem rebar of the FRP side outer
5.5
DATA FROM CASTING OF DECK CONCRETE
Following the same process as the first test specimen, concrete was cast to
create a 178 mm (7 in) thick deck slab, which applied a load per unit area of 4.21
N/mm2 (88 psf). During the pouring of the concrete, the composite deck panel had
peak deflections of less than half that of the wooden shored formwork used for the
239
reinforced concrete side, 2.2 mm (0.087 in) versus 5.9 mm (0.232 in), respectively as
shown in Figure 5.22. Forty-eight hours after the completion of the concrete pour, the
deflections for the composite side versus reinforced concrete side were 1.4 mm (0.055
in) versus 5.1 mm (0.201 in) respectively.
The deflection of the composite formwork during construction was 78.4% less
than the AASHTO specified deflection limit of 10.2 mm (0.4 mm), which was
determined from the span length divided by 180. These deflection results indicate that
the FRP panels function well as formwork, exhibiting minimal deflections under
construction loading.
Conventional Formwork
Deflection (mm)
5
Plan view of deck
4
3
2
1
FRP Panel Formwork
0
0
4
6
Time (hours)
10
Figure 5.22 Comparison of formwork center deflections under construction loading
Temperature and strain data were also taken during the construction pour to
determine the temperature of the concrete during cure and the strains imposed on the
240
deck due to construction loading and shrinkage. To measure the temperature of the
deck, three thermocouples were used with one thermocouple embedded in the center
of the composite side deck concrete, another embedded in the center of the RC side
deck concrete and the final thermocouple placed adjacent to the test specimen, to
measure ambient temperatures. Figure 5.23 shows the temperatures measured by the
three thermocouples over a 72 hour period.
45
RC side
40
Plan view of deck
Temperature (Celcius)
35
30
Composite
side
25
20
15
10
Ambient
5
0
0
10
20
30
40
50
Time (hours)
60
70
80
Figure 5.23 Temperature progression over three days for freshly poured deck concrete
(Deck pour completed within first two hours from start of data acquisition)
Note that the sudden shifts in the ambient temperature directly correspond with the
operating hours of the testing facility, in which large bay doors are opened and closed
at the start and close of business hours respectively. The peak temperature in the
composite side deck slab was 37.2 degrees Celsius and was achieved approximately
11.5 hours after- the completion of the concrete deck pour, whereas the RC side deck
241
slab achieved a peak temperature of 40.9 degrees Celsius approximately 13.5 hours
after completion of the concrete deck pour.
Strain values from the strain gages located throughout the specimen changed
by less than 100 microstrains during the twenty-four hours after the deck concrete
pour. The results from temperature and strain data measured within the deck indicate
that the change in temperature due to the exothermic reaction from the concrete deck
pour has a negligible effect on the performance of the FRP panel system.
5.6
5.6.1
STATIC LOADING OF TEST SPECIMEN

Loading Protocol
Load was applied to the specimen using the same testing setup as the first test
specimen, with two 890 kN (200 kip) capacity actuators spaced 1830 mm (6 ft) apart
from each other and at the center of each deck slab. The current test specimen was
subjected to quasi static up to 70% of the load at which punching shear occurred for
the first test specimen followed by extensive cyclic loading. This section describes the
static portion of loading, whereas the details and test results from cyclic loading are
described in Section 5.7.
The quasi static portion of testing was applied via single cycles at 98 kN (22
kip), 169 kN (38 kip), 338 kN (76 kip), 489 kN (110 kip) and 578 kN (130 kip). The
high loading levels applied to the specimen during the static phase were applied in
order to establish a performance comparison between systems and ensure the system
242
exhibits a stable response even when exposed to large overloads. A summary of the
static loading protocol is shown in Table 5.5.
In addition to the loading described in Table 5.5, two additional loading steps
were applied. After the 489 kN (110 kip) load level, the unloaded specimen was
loaded to 98 kN (22 kip) then unloaded prior to continuing on to the 578 kN (130 kip)
load level. An additional static loading of the test specimen in a single cycle up to 578
kN (130 kN) was also performed after the completion of the quasi static loading phase
and immediately prior to the start of the cyclic phases.
Table 5.5 Static loading protocol for test specimen #2
5.6.2
Overall Behavior
At each load level in which a single cycle was performed, the target load was
unloading the specimen. This allowed for easier observation of any cracking or
243
damage which had occurred to the test specimen during a given load level. The first
minor cracks due to loading were observed at the 489 kN (110 kip) load level, which
was equivalent to 5.0 times the AASHTO service load. These small discontinuous
cracks were less than 0.1mm (0.004 in) width and were located on the top surface of
the deck near the center stem. Linear potentiometer and strain gage data described in
Section 5.6.8 indicate that the FRP panel and adjacent concrete remained bonded to
each other throughout the static loading phase.
The maximum midspan deflections of the deck slabs under the applied load of
578 kN (130 kip) were 3.99 mm (0.157 in) for the FRP panel reinforced slab and 3.62
mm (0.143 in) for the reinforced concrete slab. The overall load versus midspan
displacement response curves for the two slabs are shown in Figure 5.24. In order to
better compare the structural response of the two reinforcing schemes, the unloading
portion of each curve was removed and the envelope plot of load versus center
deflections is shown in Figure 5.25.
At the 98 kN (22 kip) load level, equivalent to the AASHTO HS20 wheel load
of 73 kN (16.4 kip) with an additional 33% impact factor, the vertical midspan
deflection for the FRP side slab was 0.36 mm (0.014 in) whereas the midspan
deflection for the RC side slab was 0.29 mm (0.011 in). Throughout the loading of
the test specimen, the composite side deflected slightly more at the midspan than the
reinforced concrete side, with a maximum difference between the two slabs of 0.36
mm (0.014 in), equivalent to a 9.9% greater composite side deflection, which occurred
at a the maximum applied load level of 578 kN (130 kip).
6x AASHTO
service load
600
Composite side
RC side
500
Load (kN)
400
300
200
1x AASHTO 100
service load
0
0.0
0.5
1.0
1.5
2.0
2.5
Deflection (mm)
3.0
3.5
4.0
4.5
Figure 5.24 Load versus center deflections

244
245
6x AASHTO 600
service load
500
RC side
Composite side
Load (KN)
400
300
200
1x AASHTO 100
service load
0
0.0
0.5
1.0
1.5 2.0 2.5 3.0

Deflection (mm)
3.5
4.0
4.5
Figure 5.25. Envelope plot of load versus center deflections
5.6.3
Stiffness change per increment
Progression of damage within the deck slabs as observed through degradation

of the slope of the load-deflection response is compared using effective stiffness
ratios. The effective stiffness ratio, which is defined and discussed in greater detail in
Section 3.8.3, calculates an adaptive secant stiffness using the center deflection for
each slab and normalizes this value using the adaptive secant stiffness calculated at a
baseline load. Use of effective stiffness ratio values rather than center deflection
values provides a more reliable metric for performance comparisons the two deck
slabs. The effective stiffness ratios shown in Table 5.6 and Table 5.7 highlight the
difference in the degradation trends of the load deflection slopes for the two
reinforcing systems.
246
Table 5.6 Effective stiffness ratio of FRP side slab from linear potentiometer data
Table 5.7 Effective stiffness ratio of RC side slab from linear potentiometer data
247
The degradation trends for the two slabs are most easily observed in Figure
5.26, which plots the effective stiffness values of both slabs for each load level. The
reinforced concrete side slab displayed the most significant effective stiffness
degradation at the 254 kN (57 kip) load level, where the effective stiffness ratio for the
reinforced concrete side had decreased by 24% since the start of the current loading
phase as compared to FRP panel side, which only experienced a 9% decrease in its
effective stiffness ratio.
1.0
Composite side
RC side
0.8
0.6
0.4
0.2
0.0
98
169
254 338 400

Load Level (kN)
489
578
Figure 5.26 Effective stiffness ratios of the two slabs
The FRP panel side slab maintained higher effective stiffness ratios than the
reinforced concrete side slab. Both sides showed minor decreases in the effective
stiffness ratio, with a decrease of 33% for the FRP side slab and 37% for the
248
reinforced concrete slab effective stiffness ratio at the 578 kN (130 kip) load level,
which is 5.9 times the AASHTO factored service load.
5.6.4
Deflection Profiles
levels up to maximum applied load are presented. The primary purposes for obtaining
the displacement profiles was to determine the extent to which the two reinforcement
systems were able to distribute the wheel load over a larger area in both the
longitudinal and the transverse directions, thus helping to avoid localized punching
shear failure. Unless otherwise noted, the deflection profile figures in this section all
maintain the same deflection scale, which accommodates the maximum values
observed in the specimen at the maximum load applied during this loading phase of
578 kN (130 kip) in order provide context for each deflection value in relation to the
overall deformation of the structure.
The deflection profile along the length of the deck above the central stems is
presented in Figure 5.27. This figure indicates that the deck above the center stem
deflected evenly along its length. Comparable profiles are also observed for the
deflections above the two outer stems. The small and uniform deflections observed in
the deck above the stems indicate that the structure is loaded evenly and that the
vertical deformations of the stems do not significantly affect the overall structural
response of the test specimen. The maximum deflection in the deck above the stems
at the 578 kN (130 kip) load level was 0.64 mm (0.252 in), which was approximately
249
6.2 times smaller than the corresponding midpoint deflection of the reinforced
Deflection (mm)
concrete side slab.
4.5
Max slab
+
4.0
deflection
"0"
3.5
(578 kN)
3.0
2.5
2.0
578 kN
1.5
338 kN
169kN
1.0
98 kN
0.5
0.0
-1830 -1220
-610
0
610
1220
1830
Figure 5.27 Deflection profile along length of deck above central stem
The deflection profiles along the centerline of the two slabs are shown in
Figure 5.28 and Figure 5.29. Both slabs maintained comparably shaped deflection
profiles throughout the loading range. As expected, the highest deflection values for
both slabs were observed for the linear potentiometers below the two hydraulic
actuators, with progressively smaller deflections at distances away from these
locations. Similarly, because the centerline deflection profiles shown in Figure 5.28
and Figure 5.29 were directly below the actuators, these deflection profiles contain
higher deflection values as compared to deflection profiles offset from the centerlines,
shown in Figure 5.31 and Figure 5.32.
250
4.5
Max slab
deflection
(578 kN)
4.0
3.5
3.0
+
578 kN
"0"
489 kN
Deflection (mm)
2.5
2.0
1.5
1.0
400 kN
338 kN
254kN
169 kN
98 kN
0.5
0.0
-1830
-1220
-610
0
610
1220
1830
Figure 5.28 Deflection profile along centerline of the composite side deck
4.5
4.0
Deflection (mm)
3.5
Max slab
deflection
(578 kN)
+
"0"
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-1830
578 kN
489 kN
400 kN
338 kN
254 kN
169 kN
98 kN
-1220
-610
610
1220
Figure 5.29 Deflection profile along centerline of the RC side deck
1830
251
slabs, the superimposed deflection profiles along the centerline of the decks are shown
at different load levels. The deflection profiles shown in Figure 5.30 at the load level
of 578 kN (130 kip), equivalent to 5.9 times the AASHTO service load, are
comparably shaped for the two sides, with each of the deflection values for the two
Deflection (mm)
sides similar to within 0.5 mm (0.02 in) of each other.
4.5
Max slab
4.0
deflection
3.5
(578 kN)
3.0
2.5
RC side
2.0
Composite
1.5
side
1.0
0.5
0.0
-1830 -1220
-610
+
"0"
610
1220
1830
The deflection profiles along the length of the deck, which are offset from the
centerlines of the slabs by 305 mm (12 in) towards the outer stems, are shown in
Figure 5.31 and Figure 5.32.
These figures give information regarding the
effectiveness of the reinforcement systems to distribute the actuator load to the

surrounding regions.
252
4.5
4.0
Deflection (mm)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-1830
Max slab
deflection
(578 kN)
+
"0"
578 kN
489 kN
400 kN
338 kN
254 kN
169 kN
98 kN
-1220
-610
0
610
1220
1830

(offset from the composite side centerline by 305 mm)
4.5
4.0
Deflection (mm)
3.5
Max slab
deflection
(578 kN)
+
"0"
3.0
2.5
578 kN
2.0
489 kN
400 kN
338 kN
254 kN
169 kN
98 kN
1.5
1.0
0.5
0.0
-1830
-1220
-610
0
610
1220
1830

(offset from the RC side centerline by 305 mm)
253
Throughout the loading range, the deflection profiles in the longitudinal
direction near the outer stems shown in Figure 5.31 and Figure 5.32 illustrate
comparable deflections in both slabs, with slightly larger values observed on the FRP
panel side.
The deflection profile along the transverse centerline of the test specimen is
presented in Figure 5.33. As expected, the highest deflection values were observed
near the actuator locations and the lowest deflection values were observed above each
of the three stems. At the 578 kN (130 kip) load level, equivalent to nearly six times
AASHTO service load, the composite side displayed a 3.99 mm (0.157 in) center
deflection, while the reinforced concrete side displayed a 3.63 mm (0.143 in) center
deflection. A detailed discussion of the progression of the center deflections, as they
pertain to the effective stiffness of the slabs, is provided in Section 5.6.3, for the FRP
panel reinforced slab and for the reinforced concrete slab. Both sides maintained
deflection profiles that were to each other throughout the loading range.
4.5
Max slab
deflection
(578 kN)
4.0
"0"
3.5
578 kN
Deflection (mm)
3.0
2.5
489 kN
2.0
400 kN
1.5
338 kN
254 kN
1.0
169 kN
0.5
98 kN
0.0
0
610
1220
1830
2440
3050
3660
4880
254
Figure 5.33 Deflection profile along centerline of specimen
4270
255
5.6.5
Strain Development
examination of measurements from strain gages applied to the steel reinforcement and
FRP panels within the test specimen. The precise location of all strain gages is given
in Section 5.4.2. Note that strain values reported as negative indicate compressive
strains whereas positive values indicate tensile strains.
5.6.5.1 Comparison of Strains in FRP Panel Deck and RC Deck

The load versus strain envelope plot of the bottom longitudinal strain gages
located directly below the RC side and composite side actuators is shown in Figure
5.34. The RC side strain values have been adjusted using the method described in
Section 3.8.5, which assumes a linear strain distribution within the section, in order to
determine the apparent strains at the depth within the reinforced concrete side slab that
are in line with the FRP panel strain gage as shown in Figure 3.63.
As observed in Figure 5.34, the reinforced concrete side displays higher strain
values than the strains on the FRP panel side throughout the loading range.
The
adjusted strain in the reinforced concrete side bottom longitudinal strain gage at the
maximum applied load of 578 kN (130 kip) was 1505.8 microstrains, which was 960
microstrains greater than the corresponding strain on the FRP panel longitudinal gage
of 545.8 microstrains.
The load versus strain envelope plots for the bottom transverse strain gages
indicate that the strain for the reinforced concrete side gage initially increases at a
256
slower rate than the strain for the FRP panel side gage, as shown in Figure 5.35. For
loading above approximately 250kN (56 kip), the two strain values from the two sides
increase at comparable rates.
Load (KN)
6x AASHTO 600
service load
500
CM3L
DBL17-2*
CM3L DBL17-2*
400
300
200
1x AASHTO 100
service load
0
0
500
1000
1500
2000
Figure 5.34 Load versus strain for bottom longitudinal strain gages
(Gages are located directly below the two actuators. CM3L is on the composite side
on the underside of the FRP panel; DBL17-2 is on the RC side bottom longitudinal
steel reinforcement and therefore is adjusted using Equations 3.7 and 3.8)
The strain profile for the top transverse steel reinforcement along the centerline
of the specimen is shown in Figure 5.36. As expected for a top rebar under the given
loading conditions, the strain gages located at the middle of each deck display the
highest compression strains while the gage at the center of the two stems displays the
highest levels of tensile strain. For each of the displayed load levels, the reinforced
concrete side strain gages exhibit comparable or higher values as compared to the
corresponding strain gages on the composite side of the test specimen.
257
Load (KN)
6x AASHTO 600
service load
500
DBT10-6*
CM3T
CM3T DBT10-6*
400
300
200
1x AASHTO 100
service load
0
0
500
1000
1500
2000
Figure 5.35 Load versus strain for bottom transverse strain gages
(Gages are located directly below the two actuators. DBT10-6 is on the RC side
bottom transverse steel reinforcement; CM3T is on the composite side underside of
the FRP panel and therefore is adjusted using Equations 3.7 and 3.8)
In order to better quantify the difference in response between the two sides
individual comparisons of load versus strain plots for pairs of strain gages examined.
The load versus strain responses for the top transverse strain gages located directly
below the two actuators, shown in Figure 5.37, illustrate equivalent strain values for
the two sides during initial loading. For loads above 230 kN (52 kip) until the
maximum applied load, the RC side strain gage exhibits strain values that are between
19% and 21% larger than the FRP panel side strains.
1,500
1,000
578 kN
489 kN
400 kN
338 kN
254 kN
169kN
98 kN
500
-500
-1,000
0
610
1220
1830
2440
3050
3660
4270
4880
258
Figure 5.36 Strain profile for top transverse rebar along centerline
259
Load (KN)
6x AASHTO 600
service load
500
400
DTT10-7
DTT10-3 DTT10-7
300
DTT10-3
200
1x AASHTO 100
service load
0
-1500
-1000
-500
500
Figure 5.37 Load versus strain for top transverse strain gages
(Both gages are located on the top transverse steel reinforcement directly below the
two actuators. DTT10-7 is on the RC side and DTT10-3 is on the composite side)
The load versus strain envelope plots for the top transverse strain gages located
152 mm (6 in) away from either side of the central stem and 610 mm (2 ft) away from
the centerline of the two deck slabs are shown in Figure 5.38. The two strain gages
showed equivalent strain values up to a load of approximately 338 kN (76 kip)), which
is 3.4 times the AASHTO service load. For the remainder of the loading phase, the
reinforced concrete side strain gage increased in strain values more quickly than the
FPP panels side gage such that the maximum strain in the RC side gage during this
loading phase was 150 microstrains or 42.2% greater than the equivalent strain from
the FRP panel side strain gage. The envelope plots for load versus strain of the top
transverse strain gages located directly above the outer stems, as shown in Figure 5.39,
maintain comparable strain values throughout the loading range, with less than 50
microstrains separating the two sides at any load level.
260
Load (KN)
6x AASHTO 600
service load
500
DTT10-4
DTT10-6
DTT10-4 DTT10-6
400
300
200
1x AASHTO 100
service load
0
-500
500
1000
1500
Figure 5.38 Load versus strain for top transverse strain gages near central stem
the central stem. DTT10-6 is on the RC side and DTT10-4 is on the composite side)
Load (KN)
6x AASHTO 600
service load
500
DTT10-1
DTT10-9
DTT10-9
DTT10-1
400
300
200
1x AASHTO 100
service load
0
-500
500
1000
1500
Figure 5.39 Load versus strain for top transverse strain gages above outer stems
centerline. DTT10-9 is on the RC side and DTT10-2 is on the composite side)
261
5.6.5.2 Strain Profiles along the Longitudinal Direction of the Decks

The strain profiles along the longitudinal directions for the centerlines of the
composite side and reinforced concrete side deck slabs are shown in Figure 5.40 and
Figure 5.41 respectively. These strain profiles can be used to determine the level of
stress distribution within the bridge deck due to live loading applied to the deck
surface. Note that as discussed in Section 3.8.5, the position of the strain gages within
the composite side and RC side deck of the test specimen with respect to the depth of
the slab will have an effect on the direct comparability of the absolute magnitude of
the strains. Since the FRP side does not have a bottom layer of rebar, the strain gages
for the bottom layer on the composite side are attached the underside of the FRP
panels, which is located a distance of approximately 57 mm (2.25 in) lower within the
deck slab than the gages attached to the bottom longitudinal rebar on the reinforced
concrete side.
Regardless of the magnitude of the strain values, the strain profiles of two deck
slabs can be compared in terms of profile geometry and loads at which major changes
occur. As described previously in Section 5.6.5.1, which examines the load versus
strain plots for the center strain gages, the RC side center strain values are
significantly higher than the FRP panel side center strain values throughout all loading
levels. The composite side and reinforced concrete side deck slabs shown in Figure
5.40 and Figure 5.41 both maintain large tensile values in the strain gage directly
below the actuator and compressive strains in the strain gages located greater than 610
mm (2 ft) away from the actuator.
262
1500
+
"0"
1000
500
-500
-1830
578 kN
489 kN
400 kN
338 kN
254 kN
169kN
98 kN
-1220
-610
0
610
1220
1830
Figure 5.40 Strain profile for gages attached to bottom of FRP panel along line 3
(Line of gages is located directly along the composite side centerline)
1500
+
1000
500
578 kN
489 kN
400 kN
338 kN
254 kN
"0"
169kN
0
-500
-1830
98 kN
-1220
-610
0
610
1220
Position from center of actuator(mm)
1830
Figure 5.41 Strain profile for bottom longitudinal rebar along line 17
(Rebar is offset from RC side centerline by 140 mm towards the center stem)
263
The switch in strain values from tension to compression along the RC side
strain profile indicates that the point of inflection for the RC side slab in the
longitudinal direction is less than 610 mm (2 ft) longitudinally away from the centroid
of the actuators. The FRP panel side strains switch from tension to compression at the
intersection between the FRP panels. Since the point of inflection for both decks
occurs approximately 610 mm (2 ft) away from the centroid of the actuators and this
location is near the intersection between panels on the FRP panel side deck, further
experimental work is needed that specifically addresses the transfer of stresses
between panels.
5.6.6
Strain Progression and Maximum Strain Levels
At the 1x AASHTO service load level of 98 kN (22 kip) per actuator, all
strains in the test specimen were less than 100 microstrains, with the exception of the
FRP panel side transverse gages on the top and bottom side of the center FRP panel,
directly below the actuator, which displayed values of 106 microstrains and 110
microstrains respectively. The strain profiles of nearly all strain gages within the
specimen exhibited linear increases under increasing load up until the 489 kN (110
kip) load level. The specimen is unloaded then reloaded at this load level and the
strain values before and after the unloading exhibit minor strain discontinuities
throughout the specimen. While this was the load level at which initial damage to the
concrete composite interface was observed on the first specimen, a thorough
examination of the data indicated that full composite action between the FRP panels
and the concrete was maintained for the current specimen at this load level and
264
throughout the entire static loading phase. Details regarding detection of separation
between the FRP panel and the concrete are described in Section 5.6.8.
Prior to unloading the test specimen at the 489 kN (110 kip) load level, which
is equivalent to approximately five times the AASHTO service load, the highest strain
value of 1549 microstrains was recorded from the FRP side longitudinal strain gage on
top of the center panel, directly below the FRP side actuator. The strains in the FRP
side transverse strain gage, directly below the FRP side actuator, on the top and
bottom of the center panel, were 864 microstrains and 924 microstrains respectively.
The highest strain values observed on the RC side deck at this load level were located
adjacent to the RC side actuator in the bottom layer of transverse and longitudinal
rebar read values of 812 microstrains and 805 microstrains respectively.
At the maximum applied load of 578 kN (130 kip), equivalent to nearly six
times the AASHTO service load, the maximum strain in the deck slab was 2104
microstrains for the FRP side longitudinal strain gage on top of the center panel,
directly below the FRP side actuator. The highest strain observed for the RC side was
1145 microstrains for the bottom transverse RC side strain gage at midspan located
203 mm (8 in) longitudinally away from the centerline of the RC side actuator. The
strains for the bottom transverse RC side and composite side center gages at this load
level were 559 microstrains and 1054 microstrains respectively whereas the bottom
longitudinal strains on the RC side and composite side were 950 microstrains and 546
microstrains respectively.
The strains measured in the vertical stem rebar ranged
from -243 microstrains to 262 microstrains, with the largest strain values measured in
gages located closer to the actuators and at the tops of the stem rebar.
265
5.6.7
Residual Deflections and Strains
The presence of residual deflections and residual strains observed in the test
specimen deck slab throughout the loading protocol are an indication of damage which
has occurred. Throughout the loading steps, the residual deflection accumulated for
each load level at the midspan of each slab was determined and the results are
displayed in Table 5.8. Note that the reinforced concrete side exhibited slightly higher
residual deflection values for each load level as compared the composite side values.
The total residual deflection accumulated with loading up to the 578 kN (130 kip) load
level was 1.15 mm (0.045 in) for the composite side as compared to 1.45 mm (0.057
in) for the reinforced concrete side.
Table 5.8 Residual deflections of deck slabs (from central linear potentiometers)
266
5.6.8
During the static loading of the test specimen, examination of specific strain
and deflection data was used in order to detect the presence of a separation between
the FRP panels and the concrete. A select number of linear potentiometers were
placed above and below the composite deck in the same location such that their
deflections could be compared. The difference in deflection values obtained from
pairs of these linear potentiometers was used to obtain information regarding the level
of separation that had occurred.
A differential between the top and bottom pot
readings indicates that the concrete and composite are no longer acting together as a
connected unit and are now deflecting independent of each other.
The load versus deflection envelope plots from a pair of linear potentiometers
located on the FRP panel side deck along the transverse centerline of the test specimen
on the side of the actuator closest to the central stem are shown in Figure 5.42.
Throughout the static loading phase, the largest difference between the deflections
measured from the top and bottom side of the FRP panel side deck was 0.09 mm
(0.004 in). This difference in deflection values was 44% below the threshold value
established in Section 3.8.8 for measuring a separation between the FRP panel and the
concrete.
The values from strain gauges which were attached to the composite panel, the
bottom rebar which protrudes onto the composite panel in order to satisfy the
development length required by code, and the top rebar layer which is continuous
throughout the specimen were also examined as a means of detection of the separation
between the concrete and the FRP panels.
267
Load (KN)
6x AASHTO 600
service load
500
Bottom
Top
400
300
200
1x AASHTO 100
service load
0
0.0
0.5
1.0
1.5 2.0 2.5 3.0

Deflection (mm)
3.5
4.0
4.5
of deck- located along test specimen centerline and 610 mm away from the composite
side centerline, 178 mm from the composite side edge of the center stem
The vertical difference between the strain gauge on the bottom of the FRP
plate and the strain gauge on the bottom rebar is approximately 50mm and so similar
strain values are expected, with the strain gage on the bottom of the FRP plate reading
slightly higher strains than the strain gage on the bottom rebar. Note that regions
observed are outside the point of contraflexure for the FRP panel side deck slab in the
transverse direction, therefore compressive strains are expected for the gages attached
to the FRP panel, whereas tensile strains are expected for the gages attached to the top
transverse rebar. The lack of significant discontinuities within the strain profiles
shown in Figure 5.43 and Figure 5.44 indicate that the system maintains strain
compatibility throughout the section depth. The smooth load verses strain profiles
with the top rebar in tension, the FRP panel in compression and the bottom rebar with
268
negligible strain values also indicate that the neutral axis for the deck slab is located
adjacent to the bottom rebar.
Load (KN)
6x AASHTO 600
service load
500
Plan view
FRP panel
400
300
200
Bottom
rebar
Top rebar

Top rebar
1x AASHTO 100
Bottom rebar
service load
FRP panel
0
-1500
-1000
-500
500
(located 610 mm (2 ft) away from the FRP side actuator towards the outer stem)
Load (KN)
6x AASHTO 600
service load
500
Plan view
FRP panel
400
300
200
1x AASHTO 100
service load
Bottom
rebar
Top rebar
Top rebar
Bottom rebar
FRP panel
0
-1500
-1000
-500
500
(located 610 mm (2 ft) away from the FRP side actuator towards the central stem)
269
Based on the two detection methods described above, the bond between the FRP
panels and the concrete was determined to remain fully intact throughout the static
loading phase.
5.7
5.7.1
CYCLIC LOADING OF TEST SPECIMEN

Loading Protocol
After the static loading phase was completed, cyclic loading was performed at
increasing load levels up to 5x service load for a cumulative total of 32,000 cycles.
During each cycle, the test specimen was loaded to the designated load level, then
unloaded to a nominal zero load of 4 kN (0.9 kip). The purpose of the cyclic loading
phases was to examine the extent to which the bridge deck system exhibited a stable
structural response under repeated overloads. The different loading levels applied per
actuator for the cyclic phases are multiples of the AASHTO HS20 wheel load of 73
kN (16.4 kip) with an additional fatigue factor of 15% per wheel for the fatigue limit
state [91].
The loading protocol used is shown in Table 5.9 with a graphical
representation of the loading sequence shown in Figure 5.45.

In addition to the loading described in Table 5.9, a single static load-unload
cycle up to a load of 168 kN (37.8 kip) was performed at the start of each 5,000 cycles
such that the test specimen could be checked for cracks in the loaded and unloaded
state. At each load level in which a single cycle was performed, the target load was
unloading the specimen.
270
Table 5.9 Cyclic loading protocol for test specimen #2
568
Phase I- Quasi-static
Phase V- 5P
Load (kN)
420
Phase IV- 3P
Phase III- 2.5P
252
210
168
Phase II- 2P
10,000
17000
22,000
Cumulative number of cycles
Figure 5.45 Loading phases for test specimen #2
32,000
271
5.7.2
Crack Patterns
Prior to the start of the cyclic loading, the test specimen was carefully
inspected and any cracks from concrete shrinkage and from the static loading phase of
testing were marked on the specimen. The discontinuous hairline cracks observed on
the top surface of the deck prior to the cyclic loading phase had crack widths less than
0.1 mm (0.004 in) and ran primarily in the longitudinal direction near the center stem.
At the end of 10,000 cycles at 168 kN (37.8 kip), which is twice the AASHTO
service load, approximately ten 51 mm (2 in) long cracks located above the center
stem, running in the longitudinal direction, were observed to have crack widths of
approximately 0.4 mm (0.016 in). The remainder of cracks within the specimen
varied in width from 0.1 mm (0.004 in) to 0.3 mm (0.012 in) and length from 51 mm
(2 in) to 127 mm (5 in). Throughout the remainder of the cyclic phases, only minor
discontinuous crack progression was observed, with the largest crack at the end of all
cyclic phases only 0.7 mm (0.028 in) wide.
5.7.3
Progression of Center Deflections
Center deflections were chosen as the primary indicator by which to compare

the two deck slabs in terms of structural response and progression of damage. Since
the stability of the system during the cyclic loading was the focus of this loading
phase, changes in deflection and strain values over the course of the cyclic loading
rather than their absolute magnitudes are of primary concern. The plots of load versus
center deflections during the course of the cyclic loading phases are shown for the
600
500
Load (kN)
5x AASHTO
service load 400
300
Static loading
200
1 Cycle- 2P
10,000 Cycles- 2.5P
22,000 Cycles- 5P
1x AASHTO100
service load
27,000 Cycles- 5P
32,000 Cycles- 5P
0
0
0.5
1
1.5
2
2.5
3
Center deflection of FRP side deck (mm)
3.5
4.5
272
Figure 5.46 FRP side center deflections throughout cyclic loading phases
600
500
Load (kN)
5x AASHTO
service load 400
300
Static loading
200
1 Cycle- 2P
10,000 Cycles- 2.5P
1x AASHTO100
service load
22,000 Cycles- 5P
27,000 Cycles- 5P
32,000 Cycles- 5P
0
0
0.5
1
1.5
2
2.5
Center deflection of RC side deck (mm)
3.5
4.5
273
Figure 5.47 FRP panel side center deflections throughout cyclic loading phases
274
FRP panel side and RC side in Figure 5.47 and Table 5.10 respectively. As observed
in both figures, the slopes of the load deflection curves remain relatively constant
throughout the loading phases, whereas the residual deflections gradually increase for
both sides.
Changes in the slope of the load-deflection response as measured using
effective stiffness ratios are shown in Table 5.10 and illustrated in Figure 5.48. The
effective stiffness ratio, which is defined and discussed in detail in Section 3.8.3
calculates an adaptive secant stiffness using the center deflection for each slab and
normalizes this value using the adaptive secant stiffness calculated at a baseline load.
Use of effective stiffness ratio values rather than center deflection values provides a
more reliable metric for performance comparisons of the two decks.
Table 5.10 Effective stiffness ratios of the two slabs throughout cyclic loading
275
During the first 22,000 cycles in which the test specimen was loaded up to 252
kN (56.7 kip) per actuator, equivalent to 3 times the AASHTO fatigue service load,
Table 5.10 shows the FRP panel side exhibited negligible degradation in stiffness
whereas the RC side decreased in stiffness by approximately 10%. The decrease in
stiffness for the RC side occurred primarily during the first cyclic loading phase of
10,000 cycles at twice the AASHTO fatigue service load, with a minimal change in
stiffness during the next 12,000 cycles. Throughout the course of the final cyclic
phase in which the test specimen was loaded up 420 kN (94.8 kip), equivalent to 5
times the AASHTO fatigue service load, the FRP panel side decreased in stiffness by
13% whereas the RC side decreased by 11%. After all cyclic loading phases were
completed, the final effective stiffness ratio for the FRP panel side was 0.87, which
was 10% higher than the final RC side effective stiffness ratio value of 0.79.
1.0
Composite side
RC side
0.8
0.6
0.4
0.2
0.0
1 Cycle- 2P 10,000 Cycles- 17,000 Cycles- 22,000 Cycles- 32,000 Cycles2.5P
3P
5P
5P
Figure 5.48 Effective stiffness ratios of the two slabs throughout cyclic loading
276
Figure 5.48 illustrates the trend of the effective stiffness ratios for the two slabs in
which the FRP panel side maintains higher effective stiffness ratios throughout the
course of the cyclic loading phases.
The center deflections of the two slabs under 1x AASHTO fatigue service
load, or 84 kN (18.9 kip), are shown in Figure 5.49.
While both deck slabs meet the
AASHTO defined serviceability limitation for the deflection-to-span ratio of a deck

under service loads, defined as L/800 2.29 mm (0.090 in) [91], the reinforced concrete
Deflection at 1x service load (mm)
side deck deflected over 20% greater than the FRP panel side deck.
3.0
2.5
RC side
Composite side
2.0
1.5
Composite RC side
side
1.0
0.5
0.0
3P
5P
5P
Figure 5.49 Center deflections at 1x service load throughout cyclic loading
At the start of the cyclic loading phases, the FRP panel side deflects 1.38 mm
(0.054 in) as compared to 1.18 mm (0.047 in) for the FRP panel side. By the end of
the cyclic loading phases, the difference between the RC side and the FRP panel side
277
deflections increased from 0.20 mm (0.054 in) to 0.41 mm (0.087 in), with deflections
of 2.22 mm (0.87 in) for the RC side versus 1.80 (0.071 in) for the FRP panel side.
Differences in the residual deformations between the two systems are also
significant because residual deflections accounted for 30% to 40% of the observed
service load deflections. For the current experiment, residual deflection is defined as
the center deflection of the slabs measured while the test specimen was unloaded to
the nominal zero load of 4 kN (0.9 kip). The progressions of residual deflections for
both decks throughout the static and cyclic loading phases are shown in Figure 5.50.
Residual deflections (mm)
3.0
2.5
2.0
1.5
RC side
Composite RC side
side
Composite side
1.0
0.5
0.0
3P
5P
5P
Figure 5.50 Residual deflection comparisons throughout cyclic loading
At the start of cyclic loading, the reinforced concrete side exhibited 32% larger
residual deflections than the FRP panel side, 0.95 mm (0.037 in) versus 0.72 mm
(0.028 in) for the RC side and FRP panel side respectively. This disparity in residual
278
deflections increased to 45% after 17,000 cycles. At the end of the cyclic loading
phases, the RC side exhibited a residual deflection of 1.66 mm (0.065 in) as compared
to 1.21 mm (0.048 in). The higher residual deflections for the reinforced concrete side
provide an explanation for the larger overall deflections observed on the reinforced
concrete side at peak loads as compared to the composite side, even though the
reinforced concrete side exhibits a slightly higher stiffness.
5.7.4
Deflection Profiles
The deflection distributions for the deck slab throughout the cyclic loading
phases are shown in Figure 5.51 - Figure 5.52. These figures are shown with an
applied load of twice the AASHTO fatigue service load and contain compiled data
from each of the different important loading phases.
As observed in Figure 5.51, the
center deflection for the composite side at 2x service load is initially greater than the
deflection for the reinforced concrete side. However, over the course of the cyclic
loading phases the midspan deflection of the reinforced concrete side increases beyond
the FRP panel side deflection.
Unlike the response during the static phase, the load distribution along the
transverse direction during the cyclic loading phases is more localized for the
reinforced concrete side as compared to the FRP panel side. Throughout the cyclic
phases, the composite side maintains the same dispersed deflection profile as observed
in Figure 5.28 when low level static loads were applied. It should be noted that the
composite side only exhibited the more localized deformation during the static loading
levels greater than 4x factored service load.
Composite Side
Reinforced Concrete Side
Center-line Deflections at 2x service load (mm)
Centerline
3.0
2.5
32,000 Cycles- 5P
2.0
10,000 Cycles2.5P
22,000 Cycles- 5P
1 Cycle- 2P
1.5
1.0
0.5
Static Loading
0.0
0
610
1220
1830
2440
3050
3660
4270
4880
Distance from East Edge of Deck (mm)
279
Figure 5.51 Deflection response at 2x service load along centerline during cyclic loading
280
However, the reinforced concrete side deflection response changes during the loading
phases from a dispersed profile at the start of the static phase to a localized
deformation at the end of the cyclic phase as seen in Figure 5.51. This difference in
the deflection profiles serves as an additional indicator showing the superior structural
integrity of the composite system in response to cyclic loading.
The vertical deflection response along the longitudinal direction exhibits a very
similar profile for both the FRP panel side and the reinforced concrete side as shown
in Figure 5.52 and Figure 5.53 respectively. Since the two sides responded with very
comparable profiles, a sectional analysis was necessary to get a more quantitative side
by side comparison of the change in deflection values for a given linear potentiometer.
Four different sets of linear potentiometers placed at the midpoint between stems were
used to compare the deflection profiles in the longitudinal direction for the composite
and reinforced concrete sides.
Figure 5.58 (a) shows a schematic of the linear
potentiometers used for sectioned deflection comparisons, which are shown in Figure
5.58 (b)-(e). Section M is located directly below the actuator centerline and sections
C, B, and A were located 610 mm (2 ft), 1220 mm (4 ft) and 1525 mm (5 ft) away
from the actuator centerline respectively.
From Figure 5.55 to Figure 5.58, a similar load distribution phenomenon is
observed as described for the transverse deflection profile.
The more dispersed
deflection profile observed for the composite side under low applied loads is
maintained throughout the cyclic loading phases.
281
3.0
32,000 Cycles- 5P
2.5
"0"
22,000 Cycles- 5P
Deflection (mm)
2.0
10,000 Cycles- 2.5P

1 Cycle- 2P
1.5
Static Loading
1.0
0.5
0.0
-1830 -1525 -1220 -915
-610 -305
0
305 610
Position from Center of Actuator
915
1220 1525 1830
Figure 5.52 Deflection response of FRP side at 2x service load during cyclic loading
RC side deflections at 2x service load (mm)
3.0
2.5
"0"
2.0
32,000 Cycles- 5P
1.5
22,000 Cycles- 5P
10,000 Cycles- 2.5P
1.0
0.5
1 Cycle- 2P
Static Loading
0.0
-1830 -1525 -1220 -915 -610 -305
305
610
915 1220 1525 1830
Position from Center of Actuator
Figure 5.53 Deflection response of RC side at 2x service load during cyclic loading
282
The reinforced concrete side begins with a similar dispersed deflection profile
but as the cyclic loading progresses, the deflections on the reinforced concrete side
become more localized.
This is indicative of greater, more localized damage
experienced by the reinforced concrete side as compared to the FRP formwork system.
Figure 5.54 Positioning of linear potentiometers used for deflection comparisons
3.0
Deflection (mm)
2.5
2.0
1.5
1.0
0.5
0.0
Static
1 Cycle- 2P
10,000
22,000
32,000
Loading
Cycles- 2.5P Cycles- 5P Cycles- 5P
Figure 5.55 Composite side and RC side deflections at different cyclic loading stages
( Section A-A: 1525 mm from actuator centerline)
283
3.0
Deflection (mm)
2.5
2.0
1.5
1.0
0.5
0.0
Static
1 Cycle- 2P
10,000
22,000
32,000
Loading
Cycles- 2.5P Cycles- 5P Cycles- 5P
(Section B-B: 1220 mm from actuator centerline)
3.0
Deflection (mm)
2.5
2.0
1.5
1.0
0.5
0.0
Static
Loading
1 Cycle- 2P
10,000
22,000
Cycles- 2.5P Cycles- 5P
32,000
Cycles- 5P
(Section C-C: 610 mm away from actuator centerline)
284
3.0
Deflection (mm)
2.5
2.0
1.5
1.0
0.5
0.0
Static
Loading
1 Cycle- 2P
10,000
22,000
Cycles- 2.5P Cycles- 5P
32,000
Cycles- 5P
(Section M-M: Directly below actuator actuators)
5.7.5
Summary
During the static loading phase, two deck slabs showed comparable
performance with linear responses and negligible deterioration due to loading.
Examination of the additional single cycle to 578 kN (130 kip) immediately prior to
the start of cyclic loading also corroborated that insignificant damage had accumulated
during the static loading phase. The bond between the FRP panels and the adjacent
concrete remained intact such that full composite action for the slab was maintained
throughout the static loading phase. The FRP panel side exhibited slightly higher
center deflections, however maintained a higher effective stiffness ratio at each load
level and displayed smaller residual deflections throughout the static loading. Strain
profiles in the longitudinal direction indicated that the point of inflection for both deck
slabs occurred approximately 610 mm (2 ft) away from the center of the actuators.
285
Results from the cyclic loading phases indicate that FRP panel system
exhibited better structural response to sustained cyclic loading as compared to the
reinforced concrete deck system. The FRP panel side deck showed smaller residual
deflections, maintained higher effective stiffness ratios, and deflected 19% less than
the RC side deck under 1x service loads. Overall, the FRP panel system exhibited a
more stable structural response under cyclic loading.
USE OF NEAR SURFACE MOUNTED FRP REINFORCEMENT AS A

MEANS OF RAPID BRIDGE DECK REHABILITATION
6.1
PROJECT SPECIFIC NEED FOR FRP REHABILITATION
CALTRANS is installing sound barriers along freeways to minimize the

impact of freeway noise on local residents. These sound barriers are installed on the
barrier rail located on the edge of the bridge deck as shown in Figure 6.1. These
sound barrier walls are typically of masonry construction and their addition to bridge
deck overhangs can sometimes increase the dead load of the bridge in excess of the
original design loads. Currently the entire edge region of the box girder is removed
and rebuilt using additional reinforcement. This process is labor intensive and can
significantly impact traffic.
Figure 6.1 Sound wall placed onto the overhang of a box girder bridge deck
286
287
An alternative to partial replacement of the bridge deck is the rapid
rehabilitation of the overhang using fiber reinforced polymers (FRPs). The use of
near-surface-mounted (NSM) FRP reinforcement for rehabilitation has a number of
advantages over the more common externally bonded FRP reinforcement, including
the potential for reduced surface preparation, which in turn reduces the impact to
traffic. This reduces the likelihood of debonding failures from the concrete surface due
to significantly improved anchoring ability and improved protection from mechanical
damage provided by recessing the NSM reinforcement into the concrete surface. A
discussion of different types of near surface mounted composites, previous
applications, and observed failure modes can be found in Section 2.4.
The goal of the current investigation is to provide preliminary validation of the
use of near surface mounted composites for increasing the load carrying capacity of
the overhang region for historic concrete box girder bridges. In order to achieve this
goal, the following research tasks were accomplished:
1. evaluate the load carrying capacity and structural response for an as-built
reinforced concrete bridge deck overhang without FRP strengthening
through analytical predictions and experimental studies;
2. examine different near surface mounted rehabilitation schemes that would
provide additional strength to the bridge deck overhang so that the
overhang can to safely accommodate the increased dead load from the
addition of the sound walls;
3. implement the selected NSM strengthening design into a 1.68 m (5ft 6in)
long section of reinforced concrete bridge deck overhang and provide
288
analytical predictions of performance as well as a discussion of results
from experimental testing;
4. compare the rehabilitated specimens experimental results to theoretical
predictions and to the experimental results from the testing of the as-built
reinforced concrete specimen without FRP; and
5. provide conclusions for design and use.
6.2
SPECIMEN GEOMETRY AND CONSTRUCTION
The overall test configuration used for this experimental work consists of a
reinforced concrete two-cell box girder, with a center-to-center span of 1830 mm (6 ft)
between each of the girders and a length of 3660 mm (12 ft) as shown in Figure 3.1
and Figure 3.2. The specimen deck is 178 mm (7 in) thick and the distance from the
stem wall to the edge of the overhang is 483 mm (19 in). Information relating to the
detailing and construction of the test specimen can be found in Section 3.2.
Following construction of the test specimen, the specimen was used for a
separate test series described in Chapter 4 after completion of which two 203 mm (8
in) deep cuts located 305 mm (12 in) apart from each other were created that ran
longitudinally along the entire width of the specimen (Figure 6.2). It is noted that
previous testing was restricted to loading applied at the central section of each cell and
did not involve any load application or distress to the overhang regions. The two edge
segments of the deck bounded by the longitudinal cuts were also removed as shown in
Figure 6.2. The purpose of the cuts was to allow for multiple independent tests on
sections of edge slab 1.68 m (5ft 6in) long.
289
a) Original specimen
b) Sectioned specimen, as tested
Figure 6.2 Overall dimensions of specimen
6.3
LOADING SETUP
Vertical loads were applied to the edge region of the deck slab using two
hydraulic jacks spaced 1.83 m (6 feet) apart and mounted below the strong floor of the
testing facility. The load was transferred through two 44.5 mm (1 in) diameter
threaded rods to a steel loading beam positioned 76 mm (3 in) on-center back from the
end of the overhang section of the deck. A 51 mm (2 in) thick and 152 mm (6 in)
wide elastomeric bearing pad was placed between the steel beam and the deck slab in
order to reduce stress concentrations and provide more even loading of the test
specimen as shown in Figure 6.3. The overall test setup is shown in Figure 6.4.
290
Figure 6.3 Plan view of test fixture used to load overhang
Figure 6.4 Overall test setup
6.4
AS-BUILT TEST
In order to establish a baseline for the effectiveness of the FRP repair, the test
specimen used was isolated into separate sections as described in Section 6.2 and a
portion of the concrete box girder specimen was tested as-built, without FRP
291
rehabilitation. The following section discusses the calculations, experimental setup,
loading, test observations and results from the testing of this section of as-built
reinforced concrete bridge deck.
6.4.1
Demand Calculations
The combined dead weight of a typical sound wall and traffic barrier used for
bridges in California was calculated from the CALTRANS concrete masonry sound
wall design on bridges as shown in Figure 6.5 [159]. Using this design with normal
weight grout and concrete, the gravity load per unit length for the sound wall and
traffic barrier were determined to be 13.5 kN/m (0.92 kip/ft) and 8.1 kN/m (0.56
kip/ft) respectively, for a combined weight per unit length of 21.6 kN/m (1.5 kip/ft).
The tested section of overhang was 1600 mm (5 ft 6 in) long therefore the total load
applied to the specimen from the sound wall and traffic barrier is 36.2 kN (8.1 kip).
The load was applied to the structure by two hydraulic jacks such that each jack
applied half the total loading to the overhang. In equation form, this can be expressed
as
weight wall _ per _ jack =
weight wall
2
6.1
where
weight wall = The total load applied to the overhang due to the combined weight of the
sound wall and the traffic barrier.
292
This yields a load per hydraulic jack of approximately 18 kN (4 kip) to represent the
equivalent sound wall load, which is corresponds to a distributed load of 10.7 kN/m
(0.74 kip/ft).
Figure 6.5 Standard CALTRANS masonry sound wall design [159]
293
6.4.2
Capacity Calculations
The testing method of the deck slab overhang through a loading beam across
the entire slab width can be assumed to apply a uniformly distributed load across the
width of the deck slab. This loading condition allowed the system to be treated as a
one-way slab rather than a two way slab, thereby allowing the analysis for the problem
to be simplified from a three dimensional system to two dimensional system with the
properties of the deck slab cross-section assigned to the 2D structural element. In this
section, the shear capacity of the deck slab overhang is calculated using code provided
equations, whereas the moment capacity of for the deck slab overhang is calculated
using code provided equations and a moment curvature analysis.
The shear capacity of the slab was computed according to ACI 318-08 Section
11.3 using both the general and the more detailed calculations [118]. Note that the
California Bridge Design Specifications for reinforced concrete structures used by
CALTRANS were patterned after and are in conformity with ACI Standard 318 [152].
The general calculation for shear capacity of the slab was given by the ACI 318-08
Equation 11-3 as
Vc = 2 f c' bw d
6.2
where
f c' = Concrete compressive strength (psi)
bw = Width of the concrete slab (in)
d
= Distance from the extreme compression fiber to the centroid of the tensile
reinforcement (in)
294
This equation yields a total shear capacity of 236 kN (53 kip) for the slab, which
translates to an applied force of 118 kN (26.5 kip) per hydraulic jack. The more
detailed shear capacity equation is given by ACI 318-08 equation 11-5 as
Vd
Vc = 1.9 f c' + 2500 w u bw d
Mu
6.3
where
f c'
= Concrete compressive strength (psi)
w
= Steel reinforcement ratio of the slab in the direction perpendicular to traffic
Vu
= Factored shear in the slab at the edge of the outer vertical stem
Mu
= Factored moment in the slab at the edge of the outer vertical stem
bw
= Width of the concrete slab (in)
d
= Distance from the extreme compression fiber to the centroid of the tensile
reinforcement (in)
Equation 6.3 yields a slightly more conservative total shear capacity of 233 kN (52.4
kip) for the slab, which translates to an applied force of 116 kN (26.2 kip) per
hydraulic jack.
The moment capacity of the slab was calculated as
M n = As f y d
2
where
As
= Area of steel reinforcement in the direction perpendicular to traffic flow
fy
= Yield stress of the slab steel
d
= Distance from the compression fiber to the centroid of the tensile

reinforcement
= Depth of the equivalent rectangular compression stress block
6.4
295
Using Equation 6.4, the total moment capacity was calculated to be 97.0 kN-m
(71.6 kip-ft). The equivalent force applied through the loading beam can be obtained
by dividing the moment by the distance between the applied load and the edge of the
stem, also known as the moment arm. The equivalent applied force per hydraulic jack
was 101 kN (23 kip). Since this capacity value is the lower than the computed shear
capacity, it is predicted that flexural damage will be govern the performance of the
slab.
The moment capacity of the specimen was also found from the momentcurvature response obtained by a computer program used conventionally for structural
analysis (RESPONSE 2000) to be 117.2 kN-m (85.6 kip-ft) [160]. This corresponds
to a maximum load per hydraulic jack of 122 kN (27.5 kip). The moment curvature
response of the as-built reinforced concrete deck slab is shown in Figure 6.6 below.
140
120
Moment (kN-m)
100
80
60
40
20
0
0
50
100
150
200
Curvature (10-3 rad/m)
Figure 6.6 Moment-curvature response for as-built specimen
250
296
6.4.3
Analytical Predictions
The vertical deflections of the deck slab overhang were predicted using a
piecewise linear structural analysis that employs varying sectional properties
throughout the system. The analytical method developed for deflection predictions
relies on the fundamental equation that relates the moments and the curvatures within
a section:
M
EI
6.5
where
M
E
I
= Curvature of a section
= Moment applied a section
= Youngs modulus of a section
= Moment of inertia of a section
The moment of inertia for a reinforced concrete section reduces significantly as

the concrete cracks under applied loads. The adjusted moment of inertia for a section
at any point along a moment-curvature profile can be obtained by rearranging
Equation 6.5 as shown in Equation 6.6.
I=
M
E
6.6
where
M
E
I
= Curvature of a section
= Moment applied a section
= Youngs modulus of a section
= Moment of inertia of a section
Moment-curvature data from Response 2000, which is a computer program

conventionally used for structural analysis, is used in order to obtain values for M and
297
as shown in Figure 6.7. The moment of inertia values were calculated for each
point along the moment-curvature profile.
Figure 6.7 Adjustment of moment of inertia along moment curvature profile
Next, the moment profile for the deck slab was determined by using another
computer program conventionally used for structural analysis, RISA-2D [161]. The
deck slab was modeled as a beam which was placed on three pinned supports located
at the center of each vertical stem. The deck slab was sectioned into multiple pieces
that maintain continuity throughout the member in order to allow different properties
to be assigned to each piece. Moment values at different points along the deck slab
were determined from the moment profile due to a downward vertical load applied to
the end of the overhang region of the deck slab, as shown in Figure 6.8.
298
Figure 6.8 Moment profile on deck due to a point load at the edge of the overhang
The average moment acting within each section was determined and the
moment of inertia value for each section was changed based on the moment of inertia
values from Equation 6.6 that were obtained from moment-curvature data.
The
moment of inertia value that corresponds to the same moment applied to a given
section was used as the input property for that section. After the moment of inertia
properties were adjusted for each section, the deflections of the deck slab overhang
were calculated. This procedure was repeated at different loading levels up to loads
which correspond to the ultimate moment capacity of the system. Note that this
procedure assumes zero vertical deflection for the vertical stems, therefore this
deflection prediction method provides deflection values for the overhang relative to
the adjacent vertical stem. A discussion of the deflection predictions and comparisons
with experimental results is included in Section 6.4.7.
6.4.4
Instrumentation
The total instrumentation for this experiment consisted of 16 linear

potentiometers and 2 load cells. One central row and two outer rows, each with four
299
linear potentiometers were used to measure the vertical deflection of the deck slab.
The four linear potentiometers within each row were positioned at the midspan of the
adjacent cell, above the adjacent stem, in between the stem and the loading beam, and
directly below the loading beam, as shown in Figure 6.10 (a) and (b). The deflection
of the elastomeric bearing pad was measured using four linear potentiometers, with
one linear potentiometer at each corner of the loading beam as shown in Figure 6.9.
CL
Elastomeric Bearing Pad
2438
Linear Potentiometers
152
152
1524
1676
1829
CL
(a) Instrumentation detailing for bearing pad (b) Representative linear potentiometer
Figure 6.9 Instrumentation locations used for measuring compression of bearing pad
2438
1524
CL
610
279
76
2 3 4
A
M
152
838
B
1524
1676
1829
CL
(a) Plan view of specimen
(b) Section of deck slab with linear potentiometer details
Figure 6.10 Position of linear potentiometers for measuring deflections of deck slab (Note: not to scale)
300
301
6.4.5
Loading Protocol
In addition to the test setup described in Section 3.1, a 64 mm (2 in) diameter

hole was drilled through the deck of the specimen at a distance 76 mm (3 in) by 76
mm (3 in) on center away from the corner of the deck as shown in Figure 6.11 in order
to accommodate the spacing constraints imposed by the testing setup.
Figure 6.11 Specific test setup schematic for as-built specimen
The overhang of the deck slab was tested by incrementally increasing the
hydraulic pressure supplied to the two hydraulic jacks, which loaded the overhang
through the test setup shown in Figure 6.11. Adequate time was taken between
loading levels to ensure that the hydraulic pressure had stabilized and the pressure had
equalized as much as possible between the two jacks. The load applied to the deck
slab was monotonically increased following the loading sequence shown in Table 6.1.
The load was held briefly at each load level so that observations could be made at each
stage.
302
Table 6.1 Loading protocol for as-built test specimen
6.4.6
Experimental Results
The ultimate capacity of the slab was reached at an applied load of 114 kN (26
kips) per hydraulic jack, equivalent to a uniform distributed load of 142.5 kN/m (9.8
kip/ft), which is 6.33x the nominal wall load. Note that the additional load carrying
capacity of the deck slab overhang beyond the dead load of a single sound barrier is
necessary to resist lateral loading.
As the loading of the edge of the slab was
303
increased, the top layer of transverse reinforcement above the outer edge of the stem
yielded, followed by loss of aggregate interlock resulting in failure. The deflection of
the middle of the slab directly under the loading beam when the system was loaded to
ultimate capacity was 6.36 mm (0.25 in).
As a baseline, Figure 6.12 shows the specimen prior to testing. The markings
on the top of the deck in this figure show preexisting hairline cracks in the deck.
Cracking was first observed on the top side of the deck at the 84 kN (19 kip) load per
jack and were marked on the specimen in dark blue ink. The thin cracking on the top
of the deck surface was discontinuous and approximately followed the two top
longitudinal steel reinforcement bars adjacent to the edge of the stem wall as shown in
Figure 6.13. Minor diagonal cracks along both the central and the exterior edge of the
deck slab were also observed at this load level as seen in Figure 6.14. Small diagonal
cracks initiating on the top surface of the deck observed at each end of the specimen
are shown in Figure 6.14 (a) and (b).
Figure 6.12 Deck slab prior to experimental testing
304
Figure 6.13 Initial cracking of deck slab at 84kN (19 kip) per jack- top view of deck
(a) Detail of central edge of slab
(b) Detail of exterior edge of slab
Figure 6.14 Initial cracking of deck slab at 84kN (19 kip) per jack - side view of deck
Additional opening of small cracks was observed at the load level of 102 kN
(23 kip) per jack and these cracks were marked with red ink as shown in Figure 6.15.
The cracks that followed the two top longitudinal bars opened further and became
continuous over the majority of the specimen. Cracks oriented across the width of the
specimen also formed on the top of the slab at this load level, as seen in Figure 6.15.
When the load level of 114 kN (26 kip) per jack was reached, a large diagonal
crack opened and quickly propagated, which was clearly visible on the central edge of
the slab as shown in Figure 6.16 (a). This load level was determined to be the ultimate
305
capacity of the overhang for resisting vertical loads. The cracking progressed rapidly
along the top surface of the deck as shown in Figure 6.17 and the concrete adjacent to
the loading beam settled several millimeters as seen in Figure 6.18.
After the loading
of the specimen was completed, all testing equipment and instrumentation was fully
removed and the observed cracks were marked in orange ink. The orange diagonal
cracks on the top surface of the deck face toward the hole in the deck as shown in the
upper left-hand corner of Figure 6.19.
Figure 6.15 Crack marking of deck slab at 102 kN (23 kip) per jack- top view of deck
(a) Central side of slab
(b) Exterior side of slab
Figure 6.16 Cracking observed at ultimate capacity- side view of deck
306
Figure 6.17 Cracking observed at ultimate capacity- top view of deck
Figure 6.18 Detail of cracking at ultimate capacity in central section of deck near
loading beam
Figure 6.19 Cracks observed on top of slab tested to ultimate capacity
The loose concrete was then removed in order to better observe the failure
surfaces as shown in Figure 6.20 and Figure 6.21. Increased damage was present on
the central side of the deck as compared to the exterior side. In Figure 6.21 (b), the
307
slight deformation in the rebar due to the yielding of the steel is observed. It is also
noted that the concrete remained firmly attached beyond the longitudinal rebar.
Figure 6.20 Deck slab tested to ultimate capacity after removal of loose concrete
(a) Edge of deck prior to loose concrete removal
(b) After removal
Figure 6.21 Detail of most severely damaged section
The primary variables in defining the overall structural response of the bridge
deck slab are the load per hydraulic jack at which significant damage or failure
occurred and the corresponding center deflection of the slab, directly below the
actuator. Additional instrumentation serves to add supplementary data regarding the
deformation of the specimen during testing.
As observed in Figure 6.22, the
308
deflection of the three linear potentiometers directly below the loading beam indicate
comparable deflections for lower loading levels and higher deflections with increasing
load at the central edge of the overhang, which contains linear potentiometer B4.
A4 M4
Load per hydraulic jack (kN)
120
B4
100
80
60
A4
40
M4
20
B4
0
0
4
6
8
Deflection (mm)
(a) Load versus deflection profiles
10
10
9
8
7
6
5
4
3
2
1
0
Deflection (mm)
140
A4
M4
Ultimate
B4
A4
M4
B4
24 36 48 60 72 84 96 102114
Load Level (kN/hydraulic jack)
(b) Comparison of linear potentiometers

below loading beam
Figure 6.22 Comparisons of deflections at the edge of the deck slab overhang
At the load level of 84 kN (19 kips) per hydraulic jack where cracking in the
deck was first observed, equivalent to a uniform distributed dead load of 105 kN/m
(7.2 kip/ft) or approximately 5x the nominal wall load, linear potentiometers A4 and
M4 deflected similarly while the linear potentiometer B4 exhibited a 1.3 mm (0.051
in) or 37% greater deflection value. At the load level of 102 kN (23 kip) per hydraulic
jack where the 2nd set of crack marking took place, equivalent to a uniform distributed
dead load of 126 kN/m (8.7 kip/ft) or nearly 6x the nominal wall load, the deflection at
B4 was 2.0 mm (0.078 in) or 39% greater than the other two linear potentiometers.
309
The profiles along the center of the specimen (Figure 6.24) and at both edges (Figure
6.23 and Figure 6.25) shown below exhibit similar deflection profiles and indicate that
negligible vertical deformations occur in the deck beyond the adjacent stem wall due
to edge loading of the deck slab overhang.
1524
610
279
76
A1
A2 A3 A4
Deflection (mm)
Location of linear potentiometers

Distance from edge (mm)
0 kN
24 kN
60 kN
84 kN
102 kN
114 kN
4
6
8
10
A1
A2
A3 A4
Linear Potentiometer Designations
Figure 6.23 Deflection profile along the outer edge of specimen (Line A)
1524
610
279
76
M1
M
M2 M3 M4
Deflection (mm)

0 kN
24 kN
60 kN
84 kN
102 kN
4
6
114 kN
10
M1
M2
M3 M4
Figure 6.24 Deflection profile along center of specimen (Line M)
310
B1
B2 B3 B4
Deflection (mm)

1524
610
279
76
0 kN
24 kN
60 kN
2
4
84 kN
6
102 kN
10
114 kN
B1
B2
B3 B4
Figure 6.25 Deflection profile along the central edge of specimen (Line B)
The deflection profile shown in Figure 6.26 shows comparable deflections

along the overhang at a distance midway to the adjacent stem wall. Figure 6.27
illustrates comparable deflections directly below the point of load application along
the overhang for lower load levels with less symmetric deformations observed for
higher load levels after cracking was observed throughout the specimen.
152
A3
838
1524
M3
Deflection (mm)
0
3
0 kN
24 kN
60 kN
84 kN
102 kN
114 kN
2
4
6
8
B3
3
10
B3
M3
A3
Figure 6.26 Deflections midway along overhang (Line 3)
311
152
A4
838
1524
M4
Deflection (mm)
0
4
0 kN
24 kN
60 kN
84 kN
102 kN
114 kN
B4
4
10
B4
M4
A4
Figure 6.27 Deflections directly below loading beam (Line 4)
Through a comparison of these figures, the results indicate a symmetric

structural response for load levels prior to the initial observation of cracking in the
specimen and greater deflections on one side at higher loading levels.
6.4.7
Comparison with Theory
The max moment found via moment curvature analysis of 117.2 kN-m (85.6
kip-ft) was within 6.5 % of the actual moment applied to the structure at the max
loading of 114 kN (26 kip) per hydraulic jack, which corresponds to an applied
moment of 110.0 kN-m (81.0 kip-ft). The moment capacity estimate of 97.0 kN-m
(71.6 kip-ft), determined using the ACI 318-08 prescribed equation, was off from the
experimentally determined moment capacity by 11.8%.
Figure 6.28 compares the experimentally determined deflection of the
overhang with predicted deflection values, which were adjusted to accommodate for
the deflection of the adjacent vertical stem. The analytical prediction of deflection
312
values offers close correlation with experimental results throughout the loading range.
At 114 kN (26 kip) per jack, the experimentally determined ultimate capacity of the
deck slab, the predicted deflection was 5.3 % less than the experimentally determined
deflection, with predicted and experimental deflections of 6.03 mm (0.237 in) and
6.36 mm (0.250 in) respectively.
140
Experimental results
120
Ultimate Capacity
Analytical predictions
100
Ultimate Capacity
80
60
40
20
0
0
2
4
6
Vertical Deflection of Overhang (mm)
Figure 6.28 Deflection values from experimental results with analytical predictions
6.5
REHABILITATED TEST
The following section presents NSM FRP strengthening design options for
achieving the desired capacity increase and describes the implementation, testing and
analysis of the chosen rehabilitation design.
313
6.5.1
Calculations for Potential CFRP NSM Strengthening Schemes
The increased moment demand on the test specimens deck slab due to the
addition of the sound wall is calculated and this value is used as the basis for
determining the desired capacity increase. The corresponding total area of NSM
CFRP needed to achieve the desired moment capacity increase is calculated and
design options for five different available CFRP reinforcement products are presented.
The dead weight of a typical sound wall used for bridges in California was
calculated from the CALTRANS concrete masonry sound wall design on bridges
[159]. Using this design with normal weight concrete, the gravity load per unit length
for the sound wall was determined to be 13.5 kN/m (0.92 kip/ft). Note that the weight
of the traffic barrier is not included as part of the increased moment demand
calculation because it is assumed that the weight of the traffic barrier was already
accounted for in the original design of the deck slab overhang. The tested section of
overhang was 1600 mm (5 ft 6 in) long therefore the total load applied to the specimen
from the sound wall is 22.6 kN (5.08 kip). The equivalent moment applied to the
structure due to this dead load can be obtained by multiplying the total load applied by
the distance between the applied load and the edge of the stem, also known as the
moment arm. The equivalent additional moment demand due to the sound wall was
found to be 10.91 kN-m (8.05 kip-ft).
A successful repair would strengthen the overhang to accommodate this
increased moment demand with a reasonable safety margin. For initial calculation
purposes, a safety margin of 3 was deemed appropriate.
314
M demand _ increase = M wall 3
6.7
where
Mwall = Moment demand due to the addition of a sound wall onto the overhang
This translates to an increase in moment demand of 32.7 kN-m (24.1 kipft).
Therefore, the NSM flexural strengthening will be designed to increase the capacity of
the overhang by at least this value. The experimentally determined moment capacity
of the as-built reinforced concrete deck slab overhang without FRP was found to be
110 kN-m (81 kip-ft). Therefore, the new moment capacity after strengthening should
be at least 142.7 kN-m (105.1 kip-ft), which corresponds to a minimum required
moment capacity increase of 29.7 percent over the capacity of the as-built specimen
without FRP.
6.5.2
Analytical Strength
The increased moment capacity due to FRP strengthening is equal to the sum
of the contribution from the tension steel (compression steel is ignored for this
calculation) and the contribution from the FRP reinforcement:
a
a
M n _ strengthened = As f y d + f A f f fe d f
2
2
where
As
fy
d
a
= Total area of tension steel in slab overhang test specimen

= Experimentally determined yield strength of steel reinforcement
= Distance from extreme compression fiber to tensile reinforcement
= Depth of concrete compression block, assuming rectangular stress
distribution
6.8
315
f
df
ffe
Ef
fe
= Additional reduction factor from ACI 440.06R (Section 9.6.1) [148]

= Distance from the compression fiber to the centroid of the FRP
= Effe Effective stress in the FRP assuming elastic behavior
= Experimentally determined modulus of /elasticity of FRP
= Effective strain in FRP reinforcement
By rearranging Equation 6.8, an expression for the area of FRP reinforcement
required in order to achieve a specified moment capacity increased can be obtained:
A f _ required
M n _ strengthened As f y d
2
=
a
f f fe d f
2
6.9
The required area of FRP obtained from this expression can be used to evaluate
the feasibility of different FRP strengthening options. Note that the area of FRP
required is the total area needed for the specimen overhang and thus must be
distributed along the width of the slab overhang.
6.5.3
Options for Rehabilitation
The seven product options evaluated for this rehabilitation design were
different sizes of SIKAs pultruded carbon fiber CarboDur rods and strips as well as
Hughes Brothers pultruded carbon fiber Aslan 500 rectangular bars [162-164]. The
physical properties of each option are provided in Table 6.2 for reference.
The number of strips required to attain the desired moment capacity increase
was calculated for each of the seven potential options using calculations described in
the previous section (Equation 6.9) and the results are shown in Figure 6.29. For
316
calculation of the effective stress in the FRP, ffe, a strain of fe =0.65% was assumed
based on design recommendations for FRP post-strengthening of reinforced concrete
slabs [165, 166]. The tensile modulus for each of the different FRP reinforcement
options was obtained from manufacturer reported data. Since the FRP reinforcement
type had not been selected yet, the distance from the compression fiber to the centroid
of the FRP, df, was assumed to be the full depth of the slab. Note that this assumption
will slightly overestimate the moment capacity contribution from the FRP because for
NSMR applications, the reinforcement is located slightly below the surface of the
structure. Assuming that the centroid of the FRP reinforcement is below the surface of
the structure by a distance of between 2 mm (0.079 in) and 10 mm (0.393 in) the
calculations would have overestimated the moment capacity increase due to the FRP
reinforcement by between 1% and 6%.
Spacing requirements were also considered in the calculations performed for
each FRP strengthening option. The maximum spacing recommendations provided by
the manufacturer state that on center spacing should be limited to no more than the
lesser of 0.2 times the span length (L) or five times the slab thickness (h):
smax = min (0.2 L, 5h )
6.10
Note that the span for cantilever is taken as twice the distance to the support.
This spacing limit yields a maximum recommended spacing of 203 mm (8 in). Table
6.3 below shows the number of units needed as well as the theoretical moment
capacity increase for each type of CFRP reinforcement. As observed in Table 6.3,
spacing limitations govern rather than actual strength requirement limitations. Since
317
all seven of the design options are able to achieve the increased capacity requirements,
other aspects such as cost and constructability are now used to select the FRP
reinforcement system.
One notable difference between the installation of CFRP strips as opposed to
rods is the required depth of grooves cut into the deck. The 6.4 mm (1/4 in) diameter
rods require 12.7 mm (1/2 in) deep slots and the 9.5 mm (3/8 in) rods require 15.9 mm
(5/8 in) deep rods, while the strips only require a 4 mm (0.16 in) deep groove. From a
construction viewpoint, strips as opposed to rods are far easier to implement due to
required groove depth.
Table 6.2 Physical properties of pultruded CFRP strengthening product options [162-164]
318
319
The lower modulus of the CFRP tape of 124 GPa (18.0 Msi) versus that of the
CFRP strips, 165 GPa (23.9 Msi), resulted in appreciably greater material usage for
comparable strengthening schemes. As a comparison, the S512 CFRP strip has an
estimated moment capacity increase of 81%, whereas the #3 size CFRP tape has an
estimated moment capacity increase of only 71 % and requires an additional 19% of
material above that used for the strip to achieve this increase.
Based on guidelines, material cost considerations, and the insight that in field
determination of concrete cover would be a concern because nondestructive
determination of cover depth over large areas is considered time consuming [167], the
CarboDur strips were chosen.
Because the smallest size strip far exceeded the
required moment capacity, the CarboDur S512 strips were selected, which have a 50
mm (2 in) width. The spacing was set at 203 mm (8 in) on center for the width of the
test specimen such that nine total CFRP strips were used as shown in Figure 6.29. The
bars were extended past the point of inflection to achieve a necessary development
length of 300 mm (11.8 in).
Previously tested as-built

specimen without FRP
Current specimen to be strengthened using

NSM CFRP strips
CFRP strips spaced at 203

mm (8 in) o. c.
Ld
Point of inflection
Figure 6.29 Plan view of deck illustrating chosen CFRP strengthening scheme
320
Table 6.3 Different NSM FRP strengthening options
321
322
6.5.4
Rehabilitation Construction
The following section details the implementation of the NSMR strengthening

scheme chosen in Section 5.2. Nine (9) rectangular groves spaced at 203 mm (8 in)
o.c. were cut in the top deck of the test specimen with dimensional tolerances of 70
mm - 76 mm (2 in - 3 in) for the width and 6 mm - 13 mm ( in to in) for depth.
The grooves were each 2.74 m (8 ft) long and the cut grooves are shown in Figure
6.30. After the grooves were cut to the proper dimensions, the surface was roughened
to achieve the minimum required concrete surface profile (CSP) of 3 as defined by the
ICRI surface profile guidelines [168].
Figure 6.30 Grooves cut in deck for NSM strengthening
The CarboDur S 512 carbon fiber laminate strips were cut to length and the top
and bottom surfaces were wiped clean using methyl ethyl ketone (MEK) to remove all
residual carbon dust from the surface prior to the installation of strain gages on the top
surface of the strips. An additional cleaning with MEK was performed immediately
prior to installation of the strips into the test specimen to remove any remaining
323
contaminates and surface oxidization.
A high-modulus, high-strength, structural
epoxy paste known as SikaDur 30 was used for bonding the CFRP strips to the
concrete. The structural properties of the CarboDur S 512 strips and SikaDur 30 resin
system were experimentally determined through material characterizations performed
at the University of California, San Diego within the authors research group and these
properties are shown in Table 6.4 and Table 6.5 [169].
Table 6.4 Tensile properties of SikaDur 30 resin system [169]
Table 6.5 Tensile properties of SIKA CarboDur S512 CFRP strips [169]
After the SikaDur 30 resin system was thoroughly mixed, the neat resin was
applied to each groove as a primer using a spatula to form a uniform thickness of 1.6
324
mm (1/16 in) as shown in Figure 6.31. A specialized applicator was also used to apply
a precisely controlled thickness of resin onto each of the carbon fiber strips and the
strips were carefully placed in the grooves.
Figure 6.31 Application of resin system used in grooves to bond CFRP strips to
concrete
A rubber roller was then used to properly seat each strip, using adequate
pressure to force SikaDur 30 gel out on both sides of the laminate so that the bond line
between the concrete and FRP strip does not exceed 3 mm (1/8 in) [170]. Excess gel
was carefully removed and the installed strips are shown in Figure 6.32. After the
resin system had cured for 24 hours, a low viscosity resin system, which was used for
the wear surface applied to the top of the FRP strips, was poured over the top of the
strips up to the level of the original concrete deck. The top layer of resin was mixed
with sand to allow for improved thermal compatibility with the surrounding concrete
325
and to provide a non-skid wear surface for the top of the deck. After the installation of
the NSM CFRP strengthening scheme was completed, the instrumentation was
installed and the specimen was ready for testing to determine the effectiveness of the
repair.
Figure 6.32 CFRP strips installed
In order to monitor the curing of the CarboDur 30 resin system used to attach
the CFRP strips to the deck slab, small test samples were made using resin mixed for
installation of the strips and the samples were placed adjacent to the test specimen to
ensure comparable curing conditions. These resin samples were tested daily for a
period of seven days using both dynamic mechanical thermal analysis (DMTA) and
differential scanning calorimetry (DSC) techniques. The results obtained from these
experiments regarding the glass transition temperature, Tg, of the resin system are
shown in Table 6.6 as well as Figure 6.33 and Figure 6.34. The trends from the glass
transition temperature data over time shown below indicate that the SikaDur 30 resin
system achieved nearly full cure after approximately 4-5 days.
326
Table 6.6 Tg progression of CarboDur 30 resin used in NSMR installation
Temperature (C)
66
64
62
60
58
56
54
0
1
2
3
4
5
6
Time after NSM strip installation (days)
Figure 6.33 Progression of Tg determined from DMTA tests
Temperature (C)
46
44
42
40
38
36
34
0
1
2
3
4
5
6
Time after NSM strip installation (days)
Figure 6.34 Progression of Tg determined from DSC tests
327
6.5.5
Capacity Calculations
Following the implementation of the chosen NSM CFRP strip rehabilitation

scheme,
theoretical
predictions
for
capacity
were
recalculated
using
the
experimentally determined material properties given in Table 6.5 along with an

assumed CFRP strip embedment depth of 3 mm (1/8 in) and reduced FRP strain
capacity of 0.65% as described in Section 6.5.3. The increased moment capacity
calculation due to FRP strengthening described in Section 5.1 yields a theoretical
moment capacity of 167.3 kN-m (123.4 kip-ft), which corresponds to a 52% increase
in moment capacity over the experimentally determined value for the as-built
specimen.
The moment curvature analysis performed on the FRP rehabilitated
specimen yielded a moment capacity of 185.5 kN-m (136.4 kip-ft), which corresponds
to a 69 % increase in load carrying capacity over the as-built specimen.
Predictions of deflection at the end of the deck slab overhang were performed
using the analytical method described in Section 6.4.3.
The results from these
analytical predictions are discussed in Section 6.5.9.
6.5.6
Instrumentation
The total instrumentation for this experiment consisted of 16 linear

potentiometers, 47 strain gages and 2 load cells. One central row and two outer rows,
each with four linear potentiometers were used to measure the vertical deflection of
the deck slab. The four linear potentiometers within each row were positioned at the
midspan of the adjacent cell, above the adjacent stem, in between the stem and the
loading beam, and directly below measuring compression of bearing pad.
The
328
deflection of the elastomeric bearing pad was measured using four linear
potentiometers, with one linear potentiometer at each corner of the loading beam using
the same layout as the as-built specimen shown in Figure 6.9.
All 47 strain gages were applied to the top side of the nine pultruded CFRP
strips, parallel to the direction of the fibers as shown in Figure 6.32. The two strain
gage layout patterns used on this specimen, illustrated in Figure 6.36, were applied to
alternating CFRP strips throughout the width of the specimen as shown in Figure 6.37.
2438
1524
CL
610
279
76
CL
B
1676
M
1524
A
1
838
2 3 4
152
(a) Plan view of specimen
(b) Section of deck slab with linear potentiometer details
Figure 6.35 Position of linear potentiometers for measuring deflections of deck slab
(Note: not to scale)
329
330
Pattern #1
2362
2057
1626
1168
737
483
76
Pattern #2
2057
1168
F
483
Figure 6.36 Strain gage patterns and designations
Figure 6.37 Position of strain gages attached to CFRP strips
Pattern #
1
2
1
2
1
2
1
2
A 1
331
Figure 6.38 Completed installation of CFRP strips with full instrumentation setup
Figure 6.39 Side view of completed installation of CFRP strips with full
instrumentation setup
332
6.5.7
Loading Protocol
The overhang of the deck slab was loaded using the test setup shown in Figure
6.3 and described in Section 3.1 by incrementally increasing the hydraulic pressure
supplied to the two hydraulic jacks. Adequate time was taken between loading levels
to ensure that the hydraulic pressure had stabilized and the pressure had equalized as
much as possible between the two jacks. The load applied to the deck slab was
monotonically increased following the loading sequence shown in Table 6.7. The load
was held briefly at each load level so that observations could be made.
Table 6.7 Loading protocol used for FRP rehabilitated specimen
333
334
6.5.8
Experimental Results
The ultimate capacity of the FRP rehabilitated deck slab was reached at an
applied load of
196 kN (44 kips) per hydraulic jack, equivalent to a uniform
distributed load of 245 kN/m (16.8 kip/ft), which is 11 x the nominal sound wall load.
At this load level, the deflection of the middle of the slab under the loading beam was
8.73 mm (0.34 in). The maximum strain value achieved in the CFRP strips at ultimate
capacity was 3846 microstrains. At the ultimate capacity of the specimen, debonding
of the FRP from the concrete occurred due to a tensile failure of the concrete cover
layer located between the FRP and the top layer of rebar. This loss of compatibility
within the section was quickly followed by the opening and propagation of a large
diagonal crack along the compression strut formed with the adjacent stem wall.
Cracking was observed and marked on the specimen at the four load levels of
116 kN (26 kip), 136 kN (31 kip), 160 kN (36 kip) and 184 kN (41 kip) per hydraulic
jack. The extent of visible cracking on the top and sides of the deck of the FRP
rehabilitated specimen shown in Figure 6.40 and Figure 6.41 occurred at the load level
of 184 kN (41 kip) per hydraulic jack, which was over twice the load at which
comparable cracking was observed on the as-built specimen. The comparable initial
cracking observed on the as-built specimen occurred at a load level of 84 kN (19 kip)
per hydraulic jack and is shown in Figure 6.13 and Figure 6.14. The thin cracking on
the top of the deck surface that occurred in the FRP rehabilitated specimen observed at
the load level of 184 kN (41 kip) per hydraulic jack was discontinuous and
approximately followed the top longitudinal steel reinforcement bars adjacent to the
edge of the stem wall as shown in Figure 6.40. Minor diagonal cracks along both the
335
edges of the slab, which initiated from the top surface of the deck, are shown in Figure
6.41.
Figure 6.40 Cracking of deck slab at 184 kN (41 kip) per jack- top view of deck
(a) Detail of central edge of slab
(b) Detail of exterior edge of slab
Figure 6.41 Cracking of deck slab at 184 kN (41kip) per jack - side view of deck
When the load level of 196 kN (44 kips) per hydraulic jack was reached, the
ultimate tensile strength of the top concrete cover layer was exceeded and the bond
between the FRP and the concrete was lost. This damage was quickly followed by the
opening and propagation of a large diagonal crack along the compression strut formed
336
with the adjacent stem wall shown in Figure 6.42. This load level was determined to
be the ultimate capacity of the overhang for resisting vertical loads.
(a) Central side of slab
(b) Exterior side of slab
Figure 6.42 Cracking observed at ultimate capacity- side view of deck
The top surface of the deck slab after failure of the specimen can be observed
in Figure 6.43 and the cracking due to the interfacial failure between the FRP strips
and the concrete can be observed in the in the upper left hand corner of Figure 6.43,
adjacent to the loading beam.
Figure 6.43 Cracking of deck slab at ultimate capacity- top view of deck
337
After the loading of the specimen was completed, all testing equipment and
instrumentation was fully removed to allow for easier observation of the damage
present on the specimen. Figure 6.66 shows the top view of the deck at ultimate
capacity.
Any loose concrete was removed in order to better observe the failure
surfaces, however unlike the as-built specimen in which significant loose concrete was
removed after it was tested, nearly all of the concrete remained attached to the tested
FRP rehabilitated specimen, despite the interfacial failure that occurred between the
FRP and the concrete. Note that the debonding of the CFRP strips from the concrete
occurred adjacent to where the tensile stresses on the top of the deck are maximum
while the CFRP strips remained attached for the majority of the of the slab overhang.
Figure 6.44 Side view of tested FRP rehabilitated specimen after removal of loose
concrete
338
As observed in Figure 6.45, the three linear potentiometers directly below the
loading beam maintained comparable deflections throughout the loading range applied
to the test specimen. At the failure load for the specimen, the deflections of these
three linear potentiometers were within 10% of each other which corresponds to less
220
200
180
160
140
120
100
80
60
40
20
0
10
9
8
7
6
5
4
3
2
1
0
M4
B4
Deflection (mm)
than 1 mm (0.04 in) difference in deflection values.
A4
B4
M4
A4
4
6
8
Deflection (mm)
(a) Load versus deflection profiles
10
B4
M4
A4
B4
M4
A4
24 60 84 116 160 184 196

Load per hydralulic jack (kN)
(b) Comparison of linear potentiometers

below loading beam
Figure 6.45 Comparisons of deflections at the edge of the deck slab overhang
The profiles along the center of the specimen (Figure 6.47) and at both edges
(Figure 6.46 and Figure 6.48) shown below exhibit similar deflection profiles
throughout the loading range.
339
0

24 kN
60 kN
84 kN
Deflection (mm)
2
B2 B3 B4
B1
76
279
116 kN
136 kN
160 kN
4
6
184 kN
196 kN
10
B1
B2 B3 B4
Figure 6.46 Deflection profile along the central edge of specimen (Line B)
610
1524

24 kN
60 kN
84 kN
M1
M2 M3 M4
Deflection (mm)
116 kN
136 kN
160 kN
4
6
184 kN
196 kN
76
279
10 M1
M2 M3 M4
Figure 6.47 Deflection profile along center of specimen (Line M)
610
1524

A2 A3 A4
A1
A
610
76
279
Deflection (mm)
2
4
6
8
10
24 kN
60 kN
84 kN
116 kN
136 kN
160 kN
184 kN
196 kN
A1
A2 A3 A4
Figure 6.48 Deflection profile along the outer edge of specimen (Line A)
1524
340
The deflection profiles shown in Figure 6.49 and Figure 6.50 also indicate
comparable deflections along the overhang at a distance midway to the adjacent stem
wall and directly below the point of load application respectively.
0
152
Deflection (mm)
3
B3
24 kN
60 kN
84 kN
116 kN
136 kN
160 kN
184 kN
196 kN
838
M3
1524
A3
8
3
10
A4
M4
B4
Figure 6.49 Deflections midway along overhang (Line 3)

4
838
M4
1524
A4
4
Deflection (mm)
152
B4
24 kN
60 kN
84 kN
116 kN
136 kN
160 kN
6
184 kN
196 kN
10
A3
M3
B3
Figure 6.50 Deflections directly below loading beam (Line 4)
341
The strains in the FRP strips are also examined throughout the NSM CFRP
rehabilitated specimen. The strain profiles along the edges and the middle of the
specimen are shown in Figure 6.51, Figure 6.53, and Figure 6.52 respectively. These
strain profiles indicate that the maximum strain in the CFRP strips occurs directly
above the edge of the stem wall adjacent to the deck slab overhang, referred to with
the designation, line B. At ultimate capacity, the maximum strain in the specimen
of 3846 microstrains occurs in strain gage 4B, which is located in line B near the
middle of the specimen overhang.
The strains drop off sharply for distances further away from the end of the
overhang, with the majority of the strain gages on the opposite side of the stem wall
(line C) exhibiting less than a third of the strain values shown in line B.
Location of strain gages

(See Figure 6.37 for detailed schematic)
Strains (microstrains)
4000
3500
3000
9G 9F 9E 9D 9C 9B
9
2500
9
2000
1500
1000
500
0
(Note: Vertical gridlines in figure are
positioned at a spacing of 304.8 mm (1 ft)) -500
196 kN
184 kN
160 kN
136 kN
116 kN
84 kN
60 kN
24 kN
9G 9F 9E
9D 9C 9B
Figure 6.51 Strain profile along the central edge of specimen (Line 9)
342
5G 5F 5E 5D
5C 5B
5A


positioned at a spacing of 304.8 mm (1 ft))
4000
3500
3000
2500
2000
1500
1000
500
0
-500
196 kN
184 kN
160 kN
136 kN
116 kN
84 kN
60 kN
24 kN
5G 5F 5E
5D 5C 5B 5A
Figure 6.52 Strain profile along the middle of specimen (Line 5)
1G 1F 1E 1D
1C 1B
1A


4000
3500
3000
2500
2000
1500
1000
500
0
-500
196 kN
184 kN
160 kN
136 kN
116 kN
84 kN
60 kN
24 kN
1G 1F 1E
1D 1C 1B 1A
Figure 6.53 Strain profile along the outer edge of specimen (Line 1)
The sharp drop in strain values at distances away from the adjacent stem wall and the
insignificant strains within these regions indicate that the significantly shorter lengths
of CFRP strips could be used to optimize material usage and improve constructability
without affecting load transfer and the overall system response. The strains along line
343
B, the location where the maximum strains occur in the specimen is shown in Figure
6.54. This figure indicates that the distribution of strains was even along the specimen
until the load level of 116 kN (26 kip) per jack was reached. At this level, cracking
was first observed on the specimen and higher loading levels showed comparable but
slightly less uniform strains along the specimen. The average strain along line B in the
specimen at ultimate capacity was 3423 microstrains, whereas the minimum and
maximum strains along line B were 2943 microstrains and 3846 microstrains
respectively.
4000
B
9B
8B
7B
6B
5B
4B
3B

1B
B
3500
3000
196 kN
184 kN
160 kN
136 kN
116 kN
2500
2000
1500
84 kN
1000
60 kN
500
0
1B
3B 4B 5B 6B 7B 8B 9B
Strain Gage Designation
24 kN
Figure 6.54 Strains along the edge of the stem adjacent to the deck overhang (Line B)
Using the strain data throughout the specimen and following a procedure
described by Siem et al. [166], the shear stress between the concrete and the CFRP
strips were calculated using the following equation:
344
n,n+1 =
( n+1 n ) E L t L
(xn+1 xn )
6.11
where
(xn+1 xn ) = Distance between two strain gages
EL
= Tensile elastic modulus of the CFRP strip
tL
= Thickness of the CFRP strips
Distance from west edge (m)

2.7 2.4 2.1 1.8 1.5 1.2 0.9 0.6 0.3 0.0
5G 5F 5E 5D
5C 5B
5A

positioned at a spacing of 304.8 mm)
Stress (MPa)
2.0
1.5
1.0
0.5
0.0
196 kN
160 kN
116 kN
5G 5F 5E
5D 5C 5B 5A
Figure 6.55 Shear stress distribution within adhesive along the middle of specimen
(Line 5)
The calculated shear stress values within the adhesive are simply the mean value
between two strain gages, which ignore localized stress peaks and gradients.
After the testing of the NSM CFRP rehabilitated specimen was completed, the
overhang of the specimen was cut off and carefully removed from the rest of the test
specimen as shown in Figure 6.56 to allow for further examination of this critical
region. One point of interest to examine on the removed overhang was the actual
location of the CFRP strip reinforcement within the section. Figure 6.57 shows that
the actual embedment depth of the near surface mounted CFRP strips was
345
approximately 6 mm (0.25 in) and the actual thickness of the SikaDur 30 resin layer
used to bond the CFRP strips to the concrete was also approximately 6 mm (0.25
mm). An embedment depth of 6 mm (0.25 in) is reasonable for NSM applications
because it allows enough space for an adequate top surface of resin, which will serve
as environmental protection and the wear surface for the deck. However, the 6 mm
(0.25 mm) thickness of the SikaDur 30 bottom resin layer was exceeded the maximum
manufacturer recommended value of 3 mm (1/8 inch).
432 mm (17 in)
Figure 6.56 Location of cut for removal of FRP strengthened overhang
Figure 6.57 Detail showing actual location of reinforcement
346
While the current system performed very well, the use of an overly thick resin
layer could have had a negative effect on the overall structural response of the system.
Figure 6.58 shows the failure surface of an FRP strip, which has been detached from
the top surface of the deck. The center portion of the strip with the firmly attached
concrete was the region in which the interfacial failure in the concrete occurred, while
the outer sections of the strip were neatly detached from the resin system, due to the
method of removal of the strip.
Figure 6.58 Detail of failure surface of FRP strip
6.5.9
Comparison with Theory
The NSM CFRP rehabilitated specimen reached ultimate capacity under an

applied load of 196 kN (44 kip) per hydraulic jack which is equivalent to an applied
moment of 189.2 kN-m (139.2 kip-ft). This improved performance corresponds to a
72% increase in ultimate capacity over the as-built specimen, which failed under an
347
applied load of 114 kN (26 kip) corresponding to an applied moment of 110.0 kN-m
(81.0 kip-ft). The ACI 440-02 calculation for externally bonded FRP reinforcement
predicted a maximum moment capacity of 166.0 kN-m (122.4 kip-ft), which
corresponds to a 51% increase in load carrying capacity over the experimentally
determined capacity of the as-built specimen.
The moment curvature analysis
predicted a maximum moment capacity of 185.5 kN-m (136.4 kip-ft), which

corresponds to a 69% increase in load carrying capacity over the as-built specimen.
The theoretical predictions and experimental results were within 12% using the ACI
440 approach and were in close agreement with only a 2% error using the moment
curvature analysis. The larger disagreement with experimental data found from the
ACI 440 moment capacity increase equation likely exists because this calculation is a
more simplified approach that does not take into account the over strength of the steel
reinforcement.
The predicted deflections for the overhang provide good correlation up until a
load of approximately 160 kN/jack and under predict deflections for higher loads.
This under-prediction is likely caused by the analytical model assumption that full
composite action is maintained for the section throughout the loading range, because
failure in the concrete directly adjacent to the CFRP strips was observed.
348
220
200
180
160
140
120
100
80
60
40
20
0
Experimental results
Analytical Prediction
Ultimate Capacity
Ultimate Capacity
2
4
6
8
10
Figure 6.59 Experimental and analytical load versus deflection plots for FRP
strengthened deck overhang
6.6
COMPARISON OF FRP STRENGTHENED AND AS-BUILT OVERHANGS
The ultimate capacity of the near surface mounted CFRP strip rehabilitated
slab was reached at an applied load of 196 kN (44 kips), equivalent to a uniform
distributed load of 245 kN/m (16.8 kip/ft), which is 11 x the nominal sound wall load.
This ultimate capacity is 78% greater than that obtained by the as-built specimen,
which occurred at 114 kN (26 kips) per hydraulic jack, equivalent to a uniform
distributed load of 142.5 kN/m (9.8 kip/ft), which is 6.33x the nominal wall load. The
center deflections under the loading beam for both specimens over the complete
loading ranges applied are compared in Figure 6.60 and Figure 6.61.
349
At the failure load level of the as-built specimen, the as-built specimen had a
center deflection under the loading beam of 6.36 mm (0.25 in) whereas the FRP
rehabilitated specimen deflected approximately half that of the as-built specimen, or
3.33 mm (0.13 in). At ultimate capacity of the FRP rehabilitated specimen, the center
deflection under the loading beam was 8.73 mm (0.34 in), which indicates the
rehabilitated specimen exhibited a 31.8% increase in deformation capacity over the asbuilt specimen.
220
As-built prediction
As-built experimental
FRP strengthened prediction
FRP strenthened experimental
200
180
160
FRP strengthened test
140
120
100
80
60
40
As-built test
20
0
0
4
6
8
10
Figure 6.60 Comparison of load versus deflection profiles for deck overhang
(Deflections are measured from the location below the center of the loading beam)
Deflection (mm)
350
10
9
8
7
6
5
4
3
2
1
0
As-built
FRP Strengthened
Ultimate
Capacity
Ultimate
Capacity
24 36 48 60 72 84 96 102 114 136 160 184 196

Load per hydralulic jack (kN)
Figure 6.61 Comparisons of deflections of the deck overhang at different load levels
(Deflections are measured from the location below the center of the loading beam)
In addition to increasing the overall load and deformation capacity of the

system, the near surface mounted CFRP strips act to increase the stiffness and improve
the stability of the system. The load versus deflection profile shown in Figure 6.60
shows the significantly increased stiffness and more linear profile for the FRP
rehabilitated specimen over the as-built specimen.
The deflection profile comparison at the 114 kN (26 kip) per jack load level
along the middle of the two slabs is shown in Figure 6.62. This figure illustrates that
for the same load level, the deflections within the FRP rehabilitated specimen are
lower than the deflections of the as-built specimen throughout the deck slab and not
just at the point of load application. The first set cracking observed on the as-built
specimen occurred at the load level of 84 kN (19 kip), whereas the first set cracking
observed on the FRP rehabilitated specimen occurred at 116 kN (26 kip), which is
351
corresponds to a 38% greater load. A comparison of the deflection profiles along the
center of the specimens at these loading levels shown in Figure 6.63 reveals nearly
identical deformations for the two specimens.

610
Deflection (mm)
M2 M3 M4
M1
76
279
1524
As-built
FRP
rehabilited
6
8
10
M1
M2
M3
Linear Potentiometer Designation
M4
Figure 6.62 Comparison of center deflection profiles (Line M) at a load level of 114
kN (26 kip) per jack

M1
M2 M3 M4
Deflection (mm)
As-built
(84 kN)
FRP rehabilited
(116 kN)
6
8
610
1524
76
279
10
M1
M2
M3 M4
Figure 6.63 Comparison of center deflection profiles (Line M) at 1st observed

cracking loads
352
The second set of cracking observed on the as-built specimen occurred at the
load level of 101 kN (23 kip), whereas the second set of cracking observed on the FRP
rehabilitated specimen occurred at 136 kN (31 kip), which is corresponds to a 35%
greater load. The deflection profile comparison at the load levels where the second set
of cracking was observed in Figure 6.67 also exhibits nearly identical deformations for
the two specimens. This indicates that while the FRP reinforcement acts to stiffen the
system and increase the load carrying capacity of the overhang region, the shape of the
deflection response profile of the system is not modified significantly with the
addition of the CFRP strips.
M1
M2 M3 M4
610
1524
76
279
Deflection (mm)

As-built
2 (101 kN)
4
FRP rehabilited
(136 kN)
6
8
10
M1
M2
M3 M4
Figure 6.64 Comparison of center deflection profiles at 2nd observed cracking loads
(Line M)
Figure 6.67 shows the side by side top decks of the two specimens after testing
has been completed and all loose concrete on the top deck removed. Details of the
critical region of the as-built and FRP rehabilitated top deck are shown in Figure 6.65
353
and Figure 6.66 respectively.
For the as-built specimen, extensive damage and
spalling of the concrete on the top of the deck slab was seen. Yielding in the transverse
steel reinforcement followed by loss of aggregate interlock, resulting in failure was
observed. For the FRP rehabilitated specimen, negligible spalled concrete and loose
concrete rubble was detected. A concrete splitting failure mode was observed in the
FRP rehabilitated specimen.
Figure 6.65 Detail of cracking observed at ultimate capacity- top view of as-built deck
Figure 6.66 Detail of cracking observed at ultimate capacity- top view of FRP
rehabilitated deck
354
Figure 6.67 Top view of deck slab tested to ultimate capacity after removal of loose
concrete
355
6.7
SUMMARY OF RESULTS
Experimental results from the testing of the rehabilitated specimen indicate

that the NSMR strengthening scheme was successful at achieving the desired load
carrying capacity increase.
The ultimate load carrying capacity of the FRP
rehabilitated specimen was 196 kN (44 kips) per hydraulic jack, which was 78%
higher that the ultimate load of the as-built specimen of 114 kN (26 kips) per hydraulic
jack. This value well exceeded the desired load capacity increase of 29.7% above the
experimentally determined capacity of the as-built specimen, exhibited a very stable
structural response and increased the deformation capacity of the system.
The
theoretical moment capacity predictions for the as-built specimen were within 11%
and 6.5% of the experimentally determined value using ACI 318 and moment
curvature analysis respectively. The theoretical moment capacity predictions for the
FRP rehabilitated specimen were within 12% of the experimental value using the
modified ACI 440-02 approach and were within 2 % using moment curvature analysis.
The NSM FRP rehabilitated specimen exhibits a variety of structural
performance improvements over the as-built specimen including increased ultimate
load carrying capacity, enhanced deformation capacity and more stable overall
structural performance.
Design options for the near surface mounted CFRP
strengthening schemes allow for great flexibility in terms of tailoring the

reinforcement parameters for specific applications. With consideration of the minimal
disruption to traffic flow and ease of installation, this system is a viable and very
attractive rehabilitation option for bridge deck slab overhangs.
7
7.1
CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK

SUMMARY AND CONCLUSIONS
As the nations inventory of concrete box girder bridges deteriorate over time
and more bridges are in need of repair or replacement, the development of improved
rehabilitation techniques becomes increasingly important.
Rapid rehabilitation
techniques that use FRP materials can provide significant advantages over use of
conventional materials in terms of potential for enhanced durability, increased
construction schedules and optimized material usage.
The current research has focused primarily on two aspects related to the use of
FRP composites for rapid rehabilitation of reinforced concrete box girder bridges. The
development, assessment and validation of a FRP stay-in-place combined formwork
and reinforcement system as part of a structural system for the deck of a reinforced
concrete box girder bridge were addressed in this research. The examination of rapid
rehabilitation technique of minimally invasive flexural strengthening of box girder
bridge overhangs using NSM FRP reinforcement placed into shallow groves on the
top of the bridge deck, was also addressed. The primary findings of these research
endeavors as well as recommendations for future research are presented in this section.
7.1.1
FRP Panel SIP Structural Formwork System for Box Girder Bridges
The developed hybrid FRP-concrete bridge deck system provides an

alternative rehabilitation and replacement option for deteriorated bridge decks. The
stiffened FRP structural formwork panels serve as positive moment reinforcement in
356
357
place of corrosion prone steel as well as stay-in-place formwork for the bridge deck of
a concrete box girder bridge. By combining the installation of formwork and the
lower level of reinforcement into a single construction step, this system improves
construction efficiency and speed, resulting in significant labor savings and
accelerated construction schedules.
The experimental investigation of two full scale two-cell box girder bridge
specimens provided the current research with enhanced relevance for addressing
rehabilitation needs facing current the nations aging box girder bridges. Data from
the casting of the concrete deck showed that the change in temperature due to the
casting of the concrete had minimal impact on the performance of the system and the
deflection of the FRP panels was within allowable limits. The FRP panel bridge deck
system has significant reserve strengths, with experimentally determined safety limit
state for the system occurring at approximately nine times AASHTO factored service
load. The observed failure mode of the system was crushing of the concrete in
compression, which is the preferred failure mode for this type of hybrid FRP-concrete
system, rather than the catastrophic failure that accompanies FRP rupture. The system
exhibited a stable post-failure structural response under cyclic loading of over twice
the AASHTO factored service load.
Fundamental comparisons between a conventional steel reinforced concrete
system and the composite SIP structural formwork panel system, showed comparable
performance in terms of numerous metrics including deflection distributions,
distribution widths and strain profiles. The FRP panel system exhibited superior
performance in comparison with the RC system in terms of reduced damage
358
accumulation throughout the application of 32,000 cycles at loads up to 5x the
AASHTO cyclic load.
Table 7.1 provides a brief summary of the relative
performance of the FRP panel bridge deck system and the reinforced concrete bridge
deck system.
Table 7.1 Performance comparison between FRP panel system and RC system
The resin injection repair technique, utilized on the post-failure test specimen
to reinstate load transfer between the concrete and the composite, was successful at
improving the structural response of the system and the repaired system was able to
perform comparably to the undamaged test specimen, with superior integrity of the
concrete-FRP panel and response to cyclic loading.
359
7.1.2
Rehabilitation of Bridge Decks Using Near Surface Mounted FRP
Near surface mounted FRP composites have significant advantages over

surface mounted composites for numerous reasons including improved structural
efficiency, reduction in necessary surface preparation, better protection for the FRP
reinforcement from the surrounding environment, and decreased likelihood of
exhibiting debonding failure modes as discussed in the literature review conducted,
which is summarized in Section 2.4. Rectangular FRP strips were also shown to be
superior as compared to square or round FRP bars in terms of improved material
efficiency, smaller groove dimension needed, better three-dimensional distribution of
bond stresses, and larger perimeter to cross-sectional area ratios.
Different NSM FRP designs were evaluated as options for flexural
strengthening of a reinforced concrete bridge deck overhang.
The strengthening
scheme which most efficiently satisfied necessary strength requirements and design
guidelines while taking construction considerations into account was used to
strengthen a 1.68 m (5 ft 6 in) long section of bridge deck overhang. summarizes the
different FRP reinforcement systems evaluated as potential strengthening options and
illustrates that the chosen system, 50 mm (2 in) wide strips, easily satisfies the desired
capacity increase requirement of 29% greater moment capacity as compared to the asbuilt overhang. The chosen strengthening scheme also has 18% greater structural
efficiency than comparable strengthening schemes which use bars or rods.
A full cure of the resin system used to attach the FRP reinforcement was
determined through monitoring the progression of the resin systems glass transition
temp temperature to be achieved after approximately 4-5 days. Results from the NSM
360
FRP strengthened overhang showed a 78% increase in ultimate load carrying capacity
as well as a 32% increase in deformation capacity as compared to the as-built
specimen without FRP.
7.2
7.2.1
CONCLUSIONS
Based on this research, the following primary conclusions are summarized:

1) The assessed FRP panel bridge deck system validates the efficacy of
hybridization with optimized material usage;
2) The modular prefabricated panels and their role as stay-in-place structural
formwork enable rapid placement and significant labor savings, with a
straightforward construction process similar to a conventional reinforced
concrete bridge deck;
3) The FRP panel system has applicability for repair or replacement of bridge
decks between girders, either for small sections of a bridge deck or for the
entire bridge. The FRP panel system also indicates potential for new
construction of decks for box girder bridges or other slab-on-girder bridge
systems;
4) Concrete shear damage in the bridge deck system can be mitigated by the
presence of the FRP panels such that the punching shear failure mode,
which normally results in a catastrophic failure for a conventional
reinforced concrete deck, is avoided;
361
5) The service load the distribution width of the FRP panel system is
comparable to the reinforced concrete bridge deck system;
6) The resin injection system employed was shown be a simple and very
effective repair method for reinstating load transfer between the FRP panel
and the concrete; and
7) The results from this research indicate that the developed FRP structural
formwork system is a viable alternative to a conventional steel reinforced
concrete bridge deck;
7.2.2
Based on this research, the following primary conclusions are summarized:

1) The use of near surface mounted FRP reinforcement offers an effective
and minimally invasive rapid rehabilitation technique for strengthening
the negative moment regions of a bridge deck;
2) The addition of NSM strengthening can significantly increase both the
load carrying capacity and the deformation capacity of a concrete
bridge deck overhang;
3) The developed analytical methods provided an effective means for
predicting load carrying capacity of the NSM strengthened overhang
region as well as the deflections and strains in the system at different
loading levels. The analytical methods can be used as part of the
design process for future NSM strengthening projects;
362
4) Strengthening of a bridge deck overhang using NSM strips can be
feasibly achieved with partial closure of the bridge for only a few days,
as compared to a bridge closure for well over a month that would be
necessary for a partial rebuild of the deck using conventional reinforced
concrete construction.
5) Strains in the FRP strips are significantly reduced in regions beyond the
girder stems, which indicate that further curtailment of the FRP
reinforcement is possible in order to improve the cost effectiveness of
the system; and
7.3
7.3.1
RECOMMENDATIONS FOR FUTURE RESEARCH

There are several topics which should be addressed further through future
research.
Based on the observations regarding loads at which changes to the
composite action between the FRP panels and the concrete occurred, further
examination of the interface between the concrete and the FRP panels is
recommended. This would provide useful information on design modifications that
would ensure composite action is maintained between the two materials. Although the
analysis of the structural response of the system in terms of load transfer between
panels and from panels to surrounding concrete deck and stems was addressed in a
very preliminary form in this research, further experimental and analytical
investigations which apply loading at different positions in relation to the FRP deck
363
panels would be extremely useful. The development of detailed design guidelines
which address aspects specifically relevant to the design of this type of bridge deck
structure is also recommended.
7.3.2
Since the field of near surface mounted FRP reinforcement is still relatively
new, with the most significant research activity occurring during the past ten years,
there are a variety of aspects which can benefit from further research. There has been
significant experimental work on bond tests, however work is still needed to provide
improved correlations between bond testing experiments performance of flexural tests
of NSM FRP systems. Since the efficacy of NSM FRP strengthening is intrinsically
related to the ability of the adhesive to not only bond to the reinforcement to the
concrete substrate but also to enable efficient stress transfer, study also needs to be
conducted on adhesive rheology, performance characteristics, durability, and cure.
Also, since it is likely that the NSM will be covered by asphalt, the effect of heat due
to asphalt on the adhesive and bond should be studied.
The optimization of groove dimensions and spacing for NSM use through both
analytical and experimental study is recommended. A study of adhesive rheology and
bond quality as well as durability would also provide useful information for further
progression of this research. The examination of the minimum required development
length and the effect of the girder stems on the necessary development length are also
recommended. It is emphasized that the studies recommended should be conducted on
specimens of sufficient size since small scale tests would likely provide erroneous
364
results due to effects of scale and configuration. Finally, the development of design
guidelines and development of examples comparing NSM use to surface bonding
would provide significant benefit to this research field.
REFERENCES
1.
American Society of Civil Engineers., Report card for America's

infrastructure, 2009 Report card for America's infrastructure advisory council.,
Editor. 2009, American Society of Civil Engineers. pp. 160.
2.
American Association of State Highway and Transportation Officials.,

Bridging the gap: restoring and rebuilding the nation's bridges. 2008,
American Association of State and Highway Transportation Officials. pp. 68.
3.
National Surface Tranportation Policy and Revenue Study Commission.,

Transportation for tomorrow, in Report of the National Surface Transportation
Policy and Revenue Study Commission. 2007.
4.
Bettigole, N.H., Replacing bridge decks. Civil Engineering, 1990. 60(9): pp.
76-77.
5.
Karbhari, V.M. and L. Zhao, Use of composites for 21st century civil
infrastructure. Computer Methods in Applied Mechanics and Engineering,
2000. 185(2-4): pp. 433-454.
6.
Bettigole, N.H. and R. Robinson, Bridge decks: design, construction,

rehabilitation, replacement. 1997, New York, New York: American Society of
Civil Engineers (ASCE) Press. pp.
7.
Kay, T., Assessment and rennovation of concrete structures. 1992, Harlow,

Essex, England: Longman Scientific & Technical. 224. pp.
8.
Menzies, T.R., Corrosion forms and control for infrastructure, ASTM STP
1137, ed. V. Chaker. 1991. pp.
9.
Danish Standards Association., Repair of concrete structures to EN 1504.

2004, Jordon Hill, Oxford., Burlington, Massachusetts Elsevier ButterworthHeinemann. pp.
10.
Xanthakos, P.P., Bridge strengthening and rehabilitation. Transportation

structures series. 1996, Upper Saddle River, New Jersey: Prentice Hall. 966.
pp.
11.
Keller, T., Use of fiber reinforced polymers in bridge construction. Structural

engineering document 7. 2003, Zurich, Switzerland: International Association
for Bridge and Structural Engineering. 131. pp.
12.
Nanni, A., Composites: coming on strong. Concrete Construction, 1999. 44(1):

pp. 120-124.
365
366
13.
Zureick, A.H., B. Shih, and E. Munley, Fiber-Reinforced Polymeric Bridge

Decks. Structural Engineering Review, 1995. 7(3): pp. 257-266.
14.
Zhou, A.X., J.T. Coleman, A.B. Temeles, J.J. Lesko, and T.E. Cousins,
Laboratory and field performance of cellular fiber-reinforced polymer
composite bridge deck systems. Journal of Composites for Construction, 2005.
9(5): pp. 458-467.
15.
Wu, H.C., G. Fu, R.F. Gibson, A. Yan, K. Warnemuende, and V. Anumandla,

Durability of FRP composite bridge deck materials under freeze-thaw and low
temperature conditions. ASCE Journal of Bridge Engineering, 2006. 11(4): pp.
443-451.
16.
Salim, H.A., J.F. Davalos, P.Z. Qiao, and S.A. Kiger, Analysis and design of
fiber reinforced plastic composite deck-and-stringer bridges. Composite
Structures, 1997. 38(1-4): pp. 295-307.
17.
Salim, H.A. and J.F. Davalos, FRP composite short-span bridges: Analysis,
design and testing. Journal of Advanced Materials, 1999. 31(1): pp. 18-26.
18.
Prachasaree, W., H.V.S. Gangarao, and V. Shekar, Performance Evaluation of

FRP Bridge Deck Under Shear Loads. Journal of Composite Materials, 2009.
43(4): pp. 377-395.
19.
Park, K.T., S.H. Kim, Y.H. Lee, and Y.K. Hwang, Pilot test on a developed
GFRP bridge deck. Composite Structures, 2005. 70(1): pp. 48-59.
20.
Liu, Z.H., P.K. Majumdar, T.E. Cousins, and J.J. Lesko, Development and
evaluation of an adhesively bonded panel-to-panel joint for a FRP bridge deck
system. Journal of Composites for Construction, 2008. 12(2): pp. 224-233.
21.
Kumar, P., K. Chandrashekhara, and A. Nanni, Structural performance of a

FRP bridge deck. Construction and Building Materials, 2004. 18(1): pp. 35-47.
22.
Keller, T. and H. Gurtler, Quasi-static and fatigue performance of a cellular

FRP bridge deck adhesively bonded to steel girders. Composite Structures,
2005. 70(4): pp. 484-496.
23.
Karbhari, V.M., Fiber reinforced composite bridge systems - transition from

the laboratory to the field. Composite Structures, 2004. 66(1-4): pp. 5-16.
24.
Jeong, J., Y.H. Lee, K.T. Park, and Y.K. Hwang, Field and laboratory
performance of a rectangular shaped glass fiber reinforced polymer deck.
Composite Structures, 2007. 81(4): pp. 622-628.
25.
Harries, K., FRP bridge decks - A maturing technology. ASCE Journal of

Bridge Engineering, 2006. 11(4): pp. 382-382.
367
26.
Busel, J.P., Engineered FRP bridge decks - Solutions for the future. SAMPE
Journal, 2002. 38(5): pp. 46-48.
27.
Aref, A.J., S. Alampalli, and Y.H. He, Performance of a fiber reinforced

polymer web core skew bridge superstructure. Part 1: field testing and finite
element simulations. Composite Structures, 2005. 69(4): pp. 491-499.
28.
Alampalli, S. and J. Kunin, Load testing of an FRP bridge deck on a truss

bridge. Applied Composite Materials, 2003. 10(2): pp. 85-102.
29.
Alnahhal, W. and A. Aref, Structural performance of hybrid fiber reinforced

polymer-concrete bridge superstructure systems. Composite Structures, 2008.
84(4): pp. 319-336.
30.
Aref, A.J., Y. Kitane, and G.C. Lee, Analysis of hybrid FRP-concrete multicell bridge superstructure. Composite Structures, 2005. 69(3): pp. 346-359.
31.
Bradberry, T.E. Concrete bridge decks reinforced with fiber-reinforced

polymer bars. in Design of Structures 2001. 2001: Trasportation Research
Board.
32.
Burgueno, R., A. Davol, L. Zhao, F. Seible, and V.M. Karbhari, Flexural

behavior of hybrid fiber-reinforced polymer/concrete beam/slab bridge
component. ACI Structural Journal, 2004. 101(2): pp. 228-236.
33.
Hejll, A., B. Taljsten, and M. Motavalli, Large scale hybrid FRP composite
girders for use in bridge structures - theory, test and field application.
Composites Part B-Engineering, 2005. 36(8): pp. 573-585.
34.
Keller, T., E. Schaumann, and T. Vallee, Flexural behavior of a hybrid FRP

and lightweight concrete sandwich bridge deck. Composites Part A-Applied
Science and Manufacturing, 2007. 38(3): pp. 879-889.
35.
Schaumann, E., T. Vallee, and T. Keller, Direct load transmission in hybrid

FRP and lightweight concrete sandwich bridge deck. Composites Part AApplied Science and Manufacturing, 2008. 39(3): pp. 478-487.
36.
Cheng, L.J. and V.M. Karbhari, New bridge systems using FRP composites
and concrete: A state-of-the-art review. Progress in Structural Engineering and
Materials, 2006. 8(4): pp. 143-154.
37.
Bank, L.C. Application of FRP Composites to Bridges in the USA. in

Proceedings of the International Colloquium on Application of FRP to
Bridges. 2006. Tokyo, Japan: Japan Society of Civil Engineers (JSCE).
38.
Bakis, C.E., L.C. Bank, V.L. Brown, E. Cosenza, J.F. Davalos, J.J. Lesko, A.
Machida, S.H. Rizkalla, and T.C. Triantafillou, Fiber-reinforced polymer
368
composites for construction-state-of-the-art review. Journal of Composites for
Construction, 2002. 6(2): pp. 73-87.
39.
Keller, T., Recent all-composite and hybrid fibre-reinforced polymer bridges

and buildings. Progress in Structural Engineering and Materials, 2001. 3(1):
pp. 132-140.
40.
Berg, A.C., Analysis of a bridge deck built on U.S. highway 151 with FRP
stay-in-place forms, FRP grids, and FRP rebars. Thesis submitted to the
Department of Civil Engineering in partial fulfillment of the requirements for
the degree of Masters of Science, 2004: pp. 289.
41.
Nystrom, H.E., S.E. Watkins, A. Nanni, and S. Murray, Financial viability of

fiber-reinforced polymer (FRP) bridges. ASCE Journal of Management in
Engineering, 2003. 19(1): pp. 2-8.
42.
Benmokrane, B., E. El-Salakawy, S. El-Gamal, and S. Goulet, Construction

and testing of an innovative concrete bridge deck totally reinforced with glass
FRP bars: Val-Alain Bridge on Highway 20 East. ASCE Journal of Bridge
Engineering, 2007. 12(5): pp. 632-645.
43.
Benmokrane, B., E. El-Salakawy, A. El-Ragaby, and T. Lackey, Designing

and testing of concrete bridge decks reinforced with glass FRP bars. ASCE
Journal of Bridge Engineering, 2006. 11(2): pp. 217-229.
44.
Capozucca, R., Analysis of the experimental flexural behaviour of a concrete

beam grid reinforced with C-FRP bars. Composite Structures, 2007. 79(4): pp.
517-526.
45.
Cosenza, E., G. Manfredi, and R. Realfonzo, Development length of FRP

straight rebars. Composites Part B-Engineering, 2002. 33(7): pp. 493-504.
46.
DeLorenzis, L.A., Nanni, A., and A. L. Tegola. Flexural and shear

strengthening of reinforced concrete structures with near-surface-mounted
FRP bars. in Proceedings of the 3rd international conference on advanced
composite materials in bridges and structures. 2000. Ottawa, Canada.
47.
El-Gamal, S., E. El-Salakawy, and B. Benmokrane, Behavior of concrete

bridge deck slabs reinforced with fiber-reinforced polymer bars under
concentrated loads. ACI Structural Journal, 2005. 102(5): pp. 727-735.
48.
El-Gamal, S., E. El-Salakawy, and B. Benmokrane, Influence of reinforcement

on the behavior of concrete bridge deck slabs reinforced with FRP bars.
Journal of Composites for Construction, 2007. 11(5): pp. 449-458.
369
49.
El-Ragaby, A., E. El-Salakawy, and B. Benmokrane, Fatigue analysis of

concrete bridge deck slabs reinforced with E-glass/vinyl ester FRP reinforcing
bars. Composites Part B-Engineering, 2007. 38(5-6): pp. 703-711.
50.
El-Ragaby, A., E. El-Salakawy, and B. Benmokrane, Fatigue life evaluation of

concrete bridge deck slabs reinforced with glass FRP composite bars. Journal
of Composites for Construction, 2007. 11(3): pp. 258-268.
51.
El-Salakawy, E., B. Benmokrane, and G. Desgagne, Fibre-reinforced polymer

composite bars for the concrete deck slab of Wotton Bridge. Canadian Journal
of Civil Engineering, 2003. 30(5): pp. 861-870.
52.
El-Salakawy, E., B. Benmokrane, A. El-Ragaby, and D. Nadeau, Field

investigation on the first bridge deck slab reinforced with glass FRP bars
constructed in Canada. Journal of Composites for Construction, 2005. 9(6):
pp. 470-479.
53.
El-Salakawy, E., R. Masmoudi, B. Benmokrane, F. Briere, and G. Desgagne,

Pendulum impacts into concrete bridge barriers reinforced with glass fibre
reinforced polymer composite bars. Canadian Journal of Civil Engineering,
2004. 31(4): pp. 539-552.
54.
Theodorakopoulos, D.D. and N. Swamy, Analytical model to predict punching

shear strength of FRP-reinforced concrete flat slabs. ACI Structural Journal,
2007. 104(3): pp. 257-266.
55.
Bank, L.C., M.G. Oliva, J.S. Russell, D.A. Jacobson, M. Conachen, B. Nelson,
and D. McMonigal, Double-layer prefabricated FRP grids for rapid bridge
deck construction: case study. Journal of Composites for Construction, 2006.
10(3): pp. 204-212.
56.
Banthia, N., M. Alasaly, and S. Ma, Behavior of Concrete Slabs Reinforced

with Fiber-Reinforced Plastic Grid. Journal of Materials in Civil Engineering,
1995. 7(4): pp. 252-257.
57.
Banthia, N., C. Yan, and K. Sakai, Impact resistance of concrete plates

reinforced with a fiber reinforced plastic grid. ACI Materials Journal, 1998.
95(1): pp. 11-18.
58.
Dieter, D.A., Experimental and analytical study of concrete bridge decks

constructed with FRP stay-in-place forms and FRP grid reinforcing, in Thesis
submitted to the Department of Civil Engineering in partial fulfillment of the
requirements for the degree of Masters of Science. 2002, University of
Wisconsin-Madison: Madison, Wisconsin. pp. 226.
370
59.
Dieter, D.A., J.S. Dietsche, L.C. Bank, M.G. Oliva, and J.S. Russell. Concrete
bridge decks constructed with fiber-reinforced polymer stay-in-place forms
and grid reinforcing. in Design of Structures. Transportation Research
Record. 2002.
60.
Jeong, S.K., S. Lee, C.H. Kim, D.M. Ok, and S.J. Yoon, Flexural behavior of
GFRP reinforced concrete members with CFRP grid shear reinforcements.
Fracture and Strength of Solids Vi, Pts 1 and 2, 2006. 306-308: pp. 1361-1366.
61.
Saafi, M., Design and fabrication of FRP grids for aerospace and civil
engineering applications. Journal of Aerospace Engineering, 2000. 13(4): pp.
144-149.
62.
Schmeckpeper, E.R. and C.H. Goodspeed, Fiber-Reinforced Plastic Grid for

Reinforced-Concrete Construction. Journal of Composite Materials, 1994.
28(14): pp. 1288-1304.
63.
Yost, J.R., C.H. Goodspeed, and E.R. Schmeckpeper, Flexural performance of

concrete beams reinforced with FRP grids. Journal of Composites for
Construction, 2001. 5(1): pp. 18-25.
64.
Zhang, B.R., R. Masmoudi, and B. Benmokrane, Behaviour of one-way

concrete slabs reinforced with CFRP grid reinforcements. Construction and
Building Materials, 2004. 18(8): pp. 625-635.
65.
Bae, H.U., Design of reinforcement-free bridge decks with wide flange

prestressed precast girders, in Dissertation submitted to the Department of
Civil Engineering in partial fulfillment of the requirements for the degree of
Doctor of Philosophy. 2008, University of Wisconsin-Madison: Madison,
Wisconsin.
66.
Bank, L.C., A.P. Malla, M.G. Oliva, J.S. Russell, A. Bentur, and A. Shapira, A
model specification for fiber reinforced non-participating permanent formwork
panels for concrete bridge deck construction. Construction and Building
Materials, 2009. 23(7): pp. 2264-2677.
67.
Berg, A.C., L.C. Bank, M.G. Oliva, and J.S. Russell, Construction and cost
analysis of an FRP reinforced concrete bridge deck. Construction and Building
Materials, 2006. 20(8): pp. 515-526.
68.
Frost, S.L., Effect of stay-in-place metal forms on performance of concrete

bridge decks, M.S. Thesis. 2006, Department of Civil and Environmental
Engineering, Brigham Young University, Provo UT. 66. pp.
371
69.
Hanus, J.P., L.C. Bank, and M.G. Oliva, Combined loading of a bridge deck
reinforced with a structural FRP stay-in-place form. Construction and
Building Materials, 2009. 23(4): pp. 1605-1619.
70.
Lopez-Anido, R., P.K. Dutta, J. Bouzon, S. Morton, B. Shahrooz, and I.E.

Harik, Fatigue evaluation of FRP-concrete bridge deck on steel girders at high
temperature. Society for the Advancement of Material and Processs
Engineering, Evolving and revolutionary technologies for the new millenium.
Vol. 44. 1999: SAMPE. pp.
71.
Bank, L.C., M.G. Oliva, H.U. Bae, J.W. Barker, and S.W. Yoo, Pultruded FRP
plank as formwork and reinforcement for concrete members. Advances in
Structural Engineering, 2007. 10(5): pp. 525-535.
72.
Cheng, L.J. and V.M. Karbhari, Design approach for a FRP structural
formwork based steel-free modular bridge system. Structural Engineering and
Mechanics, 2006. 24(5): pp. 561-584.
73.
Hall, J.E. and J.T. Mottram, Combined FRP Reinforcement and Permanent
Formwork for Concrete Members. Journal of Composites for Construction,
1998. 2(2): pp. 78.
74.
National Cooperative Highway Resarch Program., Development of design

specifications and commentary for horizontally curved concrete box-girder
bridges, NCHRP Report 620, Project 12-71. 2008: Washington, D.C.
75.
Hurd, M.K., Formwork for concrete. 7th ed. Prepared under the direction of
ACI committee 347, formwork for concrete. 2005, Farmington Hills,
Michigan: American Concrete Institute. pp.
76.
Johnston, D.W., Design and construction of concrete formwork. 2nd ed.

Concrete construction engineering handbook, ed. E.G. Nawy. 2008, Boca
Raton :: CRC Press. pp.
77.
ACI Committee 622., Formwork for concrete. American Concrete Institute

Journal Proceedings, 1961. 32: pp. 993-1040.
78.
Hanna, A.S., Concrete formwork systems. Civil and environmental engineering

series ;. Vol. 2. 1999, New York :: Marcel Dekker. pp.
79.
Harding, M.A., Deck forms for girder bridges, in The Aberdeen Group. 1994,
Publication #C940771.
80.
McClelland, R., Current practice sheet No. 12- Permanent formwork for
composite bridge decks, Concrete Bridge Development Group., Editor. 2005,
CBDG.
372
81.
American Concrete Institute., Guide to formwork for concrete, ed. ACI

Committee 347. 2004, Detroit, Mich.: American Concrete Institute. pp.
82.
Bakht, B. and A. Mufti, Stay-in-place concrete formwork for deck slabs of

girder bridges. Journal of the Institution of Engineers, 1997. 78(November):
pp. 129-135.
83.
Bakht, B. and C. Lam, Behaviour of transverse confining systems for steel-free

deck slabs. ASCE Journal of Bridge Engineering, 2000. 5(2): pp. 139-147.
84.
Mufti, A.A. and J.P. Newhook, On the use of steel-free concrete bridge decks
in continuous span bridges. Canadian Journal of Civil Engineering, 1999.
26(5): pp. 667-672.
85.
Mufti, A.A., L.G. Jaeger, B. Bakht, and L.D. Wegner, Experimental

Investigation of Fiber-Reinforced Concrete Deck Slabs without Internal Steel
Reinforcement. Canadian Journal of Civil Engineering, 1993. 20(3): pp. 398406.
86.
Newhook, J. and A. Mufti, A reinforcing steel-free concrete deck slab for the
Salmon River bridge. Concrete International, 1996. 18(6): pp. 30-34.
87.
Canadian Standards Association., FRC in Deck Slabs. Canadian highway

bridge design code, Section 16.7. 2000, Rexdale, Ontario, Canada: Canadian
Standards Association (CSA). pp.
88.
Mufti, A.A., B. Bakht, and J.P. Newhook, Precast concrete decks for slab-ongirder systems: A new approach. ACI Structural Journal, 2004. 101(3): pp.
395-402.
89.
Bakht, B., A. Mufti, and G. Tadros. Transverse confinement of deck slabs by

metallic stay-in-place forms. in Proc., 4th Structural specialty conference of
the Canadian Society of Civil Engineers. 2002. Montreal: Canadian Society for
Civil Engineering.
90.
Grace, N.F., J. Hanson, and T. R.D., Inspection and deterioration of bridge

decks constructed using stay-in-place metal forms and epoxy-coated
reinforcement. 2004, Southfield, Michigan: Department of Civil Engineering,
Lawrence Technological University. pp.
91.
American Association of State Highway and Transportation Officials.,

AASHTO LRFD bridge construction specifications. 2nd ed. 2004, Washington,
D.C.: American Association of State Highway and Transportation Officials. 1
v. (various pagings). pp.
92.
Kim, G.B., K. Pilakoutas, and P. Waldron, Thin FRP/GFRC structural

elements. Cement & Concrete Composites, 2008. 30(2): pp. 122-137.
373
93.
Kim, G.B., K. Pilakoutas, and P. Waldron, Development of thin FRP

reinforced GFRC permanent formwork systems. Construction and Building
Materials, 2008. 22(11): pp. 2250-2259.
94.
Bank, L.C., A.P. Malla, M.G. Oliva, J.S. Russell, A. Bentur, and A. Shapira,
Specification and design of fiber reinforced bridge deck forms for use on wide
flange T-girders, in Report No WHRP 07-10. 2007, University of WisconsinMadison: Madison, Wisconsin.
95.
Bakeri, P.A., Analysis and deisgn of polymer composite bridge decks. Thesis
submitted to the Department of Civil Engineering in partial fulfillment of the
requirements for the degree of Master of Science. Vol. Master of Science.
1989, Cambridge, Massachusetts: Massachusetts Institute of Technology. 85.
pp.
96.
Hillman, J.R. and T.M. Murray. Innovative lightweight floor systems for steel
framed structures. in IASBE Symposium on mixed structures , including new
materials. 1990. Brussels, Belgium: International Association for Bridge and
Structural Engineering.
97.
Schaumann, E., T. Valle, and T. Keller, Direct load transmission in hybrid

FRP and lightweight concrete sandwich bridge deck. Composites Part AApplied Science and Manufacturing, 2008. 39(3): pp. 478-487.
98.
Keller, T., E. Schaumann, and T. Valle, Flexural behavior of a hybrid FRP

and lightweight concrete sandwich bridge deck Composites Part A: Applied
Science and Manufacturing, 2006. 38(3): pp. 879.
99.
Schaumann, E., T. Valle, and T. Keller, Shear resistance of lighweight

concrete core of fiber-reinforced polymer concrete sandwich structure. ACI
Materials Journal, 2009. 106(1): pp. 33-41.
100.
Deskovic, N., T.C. Triantafillou, and U. Meier, Innovative Design of FRP

Combined with Concrete: Short-Term Behavior. Journal of Structural
Engineering, 1995. 121(7): pp. 1069-1078.
101.
Deskovic, N., U. Meier, and T.C. Triantafillou, Innovative Design of FRP

Combined with Concrete: Long-Term Behavior. Journal of Structural
Engineering, 1995. 121(7): pp. 1079-1089.
102.
Deskovic, N., Innovative design of FRP composite members combined with

concrete, in Civil Engineering. 1993, Massachusetts Institute of Technology:
Massachusetts. pp. 317.
103.
Banks, W.M. and J. Rhodes. Instability of composite channel sections. 1983.

Paisley, Scotl: Applied Science Publ.
374
104.
Holmes, M. and D.J. Just, GRP in structural engineering. 1983, London ; New
York :: Applied Science Publishers. pp.
105.
Canning, L., L. Hollaway, and A.M. Thorne, An investigation of the composite

action of an FRP/concrete prismatic beam. Construction and Building
Materials, 1999. 13(8): pp. 417-426.
106.
Hulatt, J., L. Hollaway, and A. Thorne, The use of advanced polymer

composites to form an economic structural unit. Construction and Building
Materials, 2003. 17(1): pp. 55-68.
107.
Kitane, Y., A.J. Aref, and G.C. Lee, Static and fatigue testing of hybrid fiberreinforced polymer-concrete bridge superstructure. Journal of Composites for
Construction, 2004. 8(2): pp. 182-190.
108.
Aref, A.J., S. Alampalli, and Y.H. He, Performance of a fiber reinforced

polymer web core skew bridge superstructure. Part II: failure modes and
parametric study. Composite Structures, 2005. 69(4): pp. 500-509.
109.
Alnahhal, W., A. Aref, and S. Alampalli, Composite behavior of hybrid FRPconcrete bridge decks on steel girders. Composite Structures, 2008. 84(1): pp.
29-43.
110.
Reising, R.M.W., B.M. Shahrooz, V.J. Hunt, A.R. Neumann, A.J. Helmicki,
and M. Hastak, Close look at construction issues and performance of four
fiber-reinforced polymer composite bridge decks. Journal of Composites for
Construction, 2004. 8(1): pp. 33-42.
111.
Reising, R.M.W., B.M. Shahrooz, V.J. Hunt, A.R. Neumann, and A.J.
Helmicki, Performance comparison of four fiber-reinforced polymer deck
panels. Journal of Composites for Construction, 2004. 8(3): pp. 265-274.
112.
Reising, R.M.W., B.M. Shahrooz, V.J. Hunt, M.S. Lenett, S. Christopher, A.R.
Neumann, A.J. Helmicki, R.A. Miller, S. Kondury, and S. Morton,
Performance of five-span steel bridge with fiber-reinforced polymer composite
deck panels. Design of Structures 2001, 2001(1770): pp. 113-123.
113.
Reising, R.M.W., Testing and long-term monitoring of a five-span bridge with

multiple FRP decks- performance and design issues. Dissertation submitted to
the Department of Civil and Environmental Engineering in partial fulfillment
of the requirements for the degree of Doctorate of Philosophy. 2003,
Cincinnati, Ohio: University of Cincinnati. 350. pp.
114.
Harik, I.E., P. Alagusundaramoorthy, R. Siddiqui, R. Lopez-Anido, S. Morton,

and P.K. Dutta. Testing of concrete/fiber-reinforced polymer composite deck
375
panels. in Proceedings of the fifth ASCE materials engineering conference.
1999. Cincinnati, Ohio: American Society of Civil Engineers.
115.
Alagusundaramoorthy, P., I.E. Harik, and C.C. Choo, Structural behavior of

FRP composite bridge deck panels. ASCE Journal of Bridge Engineering,
2006. 11(4): pp. 384-393.
116.
Dutta, P.K., R. Lopez-Anido, and S.C. Kwon, Fatigue durability of FRP

composite bridge decks at extreme temperatures. International Journal of
Materials & Product Technology, 2007. 28(1-2): pp. 198-216.
117.
Bank, L.C., Z. Xi, and E. Munley. Tests of full-sized pultruded FRP grating
reinforced concrete bridge decks. in ASCE materials engineering congress.
1992. Atlanta, Georgia: American Society of Civil Engineers.
118.
American Concrete Institute., Building Code Requirements for Structural

Concrete. 2008, Detroit, Mich: ACI 318-08. 430. pp.
119.
Cheng, L.J., L. Zhao, V.M. Karbhari, G.A. Hegemier, and F. Seible,

Assessment of a steel-free fiber reinforced polymer-composite modular bridge
system. Journal of Structural Engineering-ASCE, 2005. 131(3): pp. 498-506.
120.
Cheng, L., Development of a steel-free FRP-concrete slab-on-grider modular

bridge system, in Structural Engineering. 2005, University of California: La
Jolla, California. pp. 388.
121.
Cheng, L.J. and V.M. Karbhari, Fatigue behavior of a steel-free FRP-concrete

modular bridge deck system. ASCE Journal of Bridge Engineering, 2006.
11(4): pp. 474-488.
122.
Brinckerhoff, P., Bridge inspection and rehabilitation : a practical guide, ed.

L.G. Silano. 1993, New York, N.Y.: Wiley. xiii, 288 p. pp.
123.
Allen, R.T.L., S.C. Edwards, and J.D.N. Shaw, The Repair of concrete
structures. 2nd ed. 1993, London ; New York: Blackie Academic &
Professional. xii, 212 p. pp.
124.
Tonias, D.E., Bridge engineering : design, rehabilitation, and maintenance of

modern highway bridges. 1995, New York: McGraw-Hill. 470 p. pp.
125.
Mailvaganam, N.P., Repair and protection of concrete structures. 1992, Boca

Raton, Fla.: CRC Press. 473 p. pp.
126.
American Concrete Institute., Guide for the design and construction of

externally bonded FRP systems for strengthening concrete structures. ACI
manual of concrete practice, ed. A. 440.2R-08. 2008, Detroit, Mich.: American
Concrete Institute. 75. pp.
376
127.
Rizkalla, S., T. Hassan, and N. Hassan, Design recommendations for the use of
FRP for reinforcement and strengthening of concrete structures. Progress in
Structural Engineering and Materials, 2003. 5(1): pp. 16-28.
128.
Ghosh, K.K., Assessment of FRP composite strengthened reinforced concrete

bridge structures at the component and systems level through progressive
damage and Non-Destructive Evaluation (NDE). Dissertation submitted to the
Department of Structural Engineering in partial fulfillment of the requirements
for the degree of Doctorate of Philosophy. 2006, La Jolla, California: Universit
of California, San Diego. 481. pp.
129.
De Lorenzis, L. and J.G. Teng, Near-surface mounted FRP reinforcement: An

emerging technique for strengthening structures. Composites Part BEngineering, 2007. 38(2): pp. 119-143.
130.
Asplund, S.O., Strengthening bridge slabs with grouted reinforcement. Journal

of the American Concrete Institute, 1949. 45(1): pp. 397-406.
131.
Parretti, R. and A. Nanni, Strengthening of RC members using near-surface

mounted FRP composites: Design overview. Advances in Structural
Engineering, 2004. 7(6): pp. 469-483.
132.
Nanni, A., T. Alkhrdaji, M. Barker, G. Chen, R. Mayo, and X.B. Yang.

Overview of testing to failure program of a highway bridge strengthened with
FRP composites. in FRPRCS-4. 1999. Baltimore, Maryland.
133.
Nordin, H. and B. Taljsten, Concrete beams strengthened with prestressed

near surface mounted CFRP. Journal of Composites for Construction, 2006.
10(1): pp. 60-68.
134.
Szabo, Z.K. and G.L. Balazs, Near surface mounted FRP reinforcement for
strengthening of concrete structures. Periodica Polytechnica-Civil
Engineering, 2007. 51(1): pp. 33-38.
135.
Hollaway, L.C. and P.R. Head, Advanced polymer composites and polymers in
the civil infrastructure. 1 ed. 1991: Elsevier. 316. pp.
136.
Blaschko, M., Bond behaviour of CFRP strips glued into slits, in Fibrereinforced polymer reinforcement for concrete structures, FRPRCS-6, H.K.
Tan, Editor. 2003: Singapore. pp. 205-214.
137.
Teng, J.G., L. De Lorenzis, B. Wang, R. Li, T.N. Wong, and L. Lam,

Debonding failures of RIC beams strengthened with near surface mounted
CFRP strips. Journal of Composites for Construction, 2006. 10(2): pp. 92-105.
138.
Burke, P.J., Low and high temperature performance of near surface mounted
FRP strengthened concrete slabs, in Thesis submitted to the Department of
377
Civil Engineering in conformity with requirements for the degree of Master of
Science. 2008, Queen's University: Kingston, Ontario, Canada.
139.
Concrete Society., Design guidance for strengthening concrete structures

using fibre composite materials. 2nd ed. Concrete Society technical report No.
55. 2004, Camberley, Surrey, UK: Cromwell Press. 102. pp.
140.
De Lorenzis, L., A. Rizzo, and A. La Tegola, A modified pull-out test for bond
of near-surface mounted FRP rods in concrete. Composites Part BEngineering, 2002. 33(8): pp. 589-603.
141.
El-Hacha, R. and S.H. Rizkalla, Near-surface-mounted fiber-reinforced

polymer reinforcements for flexural strengthening of concrete structures. Aci
Structural Journal, 2004. 101(5): pp. 717-726.
142.
Yost, J.R., S.P. Gross, D.W. Dinehart, and J.J. Mildenberg, Flexural behavior
of concrete becams strengthened with near-surface-mounted CFRP strips. ACI
Structural Journal, 2007. 104(4): pp. 430-437.
143.
Tang, W.C., R.V. Balendran, A. Nadeem, and H.Y. Leung, Flexural

strengthening of reinforced lightweight polystyrene aggregate concrete beams
with near-surface mounted GFRP bars. Building and Environment, 2006.
41(10): pp. 1381-1393.
144.
Barros, J.A.O., S.J.E. Dias, and J.L.T. Lima, Efficacy of CFRP-based

techniques for the flexural and shear strengthening of concrete beams. Cement
& Concrete Composites, 2007. 29(3): pp. 203-217.
145.
Warren, G., Waterfront repair and upgrade, advanced

demonstration site No. 2: Pier 12, NAVSTA, San Diego
technology
in Site specific report SSR-2419-SHR. 1998, Naval facilities engineering service

center, Port Hueneme, CA: San Diego, CA.
146.
Farmer, N., Near-surface-mounted reinforcement preserves car park for

Merseyside police. Concrete, 2003. 37(9): pp. 36-38.
147.
Alkhradaji, T., A. Nanni, G. Chen, and M. Barker, Upgrading the

transportation infrastructure: solid RC decks strengthened with FRP. Concrete
International, 1999. 21(10): pp. 37-41.
148.
American Concrete Institute., Guide for the design and construction of

externally bonded FRP systems for strengthening concrete structures, ed. A.
440.06R. 2006, Detroit, Mich.: American Concrete Institute. pp.
149.
Canadian Standards Association., Canadian highway bridge design codeSection 15: Fibre-reinforced structures. 2006. 1 v. (various pagings). pp.
378
150.
Mufti, A.A., B. Bakht, N. Banthia, B. Benmokrane, G. Desgagne, R. Eden,

M.A. Erki, V. Karbhari, J. Kroman, D. Lai, A. Machida, K. Neale, G. Thadros,
and B. Taljsten, New Canadian highway bridge Design code design provisions
for fibre-reinforced structures. Canadian Journal of Civil Engineering, 2007.
34(3): pp. 267-283.
151.
American Society for Testing and Materials., Annual book of ASTM standards.
2000, Philadelphia, PA: American Society for Testing and Materials (ASTM).
pp.
152.
California Department of Transportation., Bridge Design Specifications,

Section 8- Reinforced concrete. Vol. Revision Number 32. 2004, Sacramento,
CA: California Department of Transportation. pp.
153.
MacGregor, J.G. and J.K. Wight, Reinforced concrete : mechanics and design.
4th ed. 2005, Upper Saddle River, N.J.: Prentice Hall. xx, 1132 p. pp.
154.
Nawy, E.G., A. Scanlon, and American Concrete Institute., Designing concrete

structures for serviceability and safety. 1992, Detroit: American Concrete
Institute. vi, 346 p. pp.
155.
American Concrete Institute., Building Code Requirements for Structural

Concrete. 2005, Detroit, Mich: ACI 318-05. pp.
156.
Sika Corporation., Product data sheet for

www.sikaconstruction.com. 2008: Lyndhurst, NJ.
157.
Loyola Construction Products., Simposon Epoxy-Tie Technical Data. 2008.
158.
American Concrete Institute., Anchorage to Concrete- Designing structural

connections to concrete, SCM-31 (06). 2006, Detroit, Mich.: American
Concrete Institute. pp.
159.
California Department of Transportation., Memo to designers: 22-1. Soundwall

design criteria. 2004, Sacramento, California: California Department of
Transportation. pp.
160.
Bentz, E. and M.P. Collins, RESPONSE 2000- Reinforced concrete sectional

analysis using the modified compression field theory. 2000: Toronto, Canada.
161.
Bates, B.R., RISA-2D Educational Version 1.0. 2002, Risa Technologies.
162.
Sika Corporation., Product data sheet for Sika CarboDur- carbon fiber
laminate for structural strengthening, in www.sikaconstruction.com. 2005:
Lyndhurst, NJ.
SikaDur
55
SLV,
in
379
163.
Sika Corporation., Product data sheet for Sika CarboDur rods- carbon fiber
rods for structural strengthening, in www.sikaconstruction.com. 2003:
Lyndhurst, NJ.
164.
Hughes Brothers Incorporated., Product data sheet forAslan 500- CFRP tape,
in www.hughesbros.com. 2002: Lyndhurst, NJ.
165.
Seim, W., M. Horman, V. Karbhari, and F. Seible, External FRP

poststrengthening of scaled concrete slabs. Journal of Composites for
Construction, 2001. 5(2): pp. 67-75.
166.
Seim, W., A. Vasquez, V. Karbhari, and F. Seible, Poststrengthening of

concrete slabs: Full-scale testing and design recommendations. Journal of
Structural Engineering-ASCE, 2003. 129(6): pp. 743-752.
167.
Sohanghpurwala, A.A., Manual on service life prediction of corrosiondamaged reinforced concrete bridge superstructure elements, in National
Cooperative Highway Resarch Program., . 2006: Washington, D.C. pp. 59.
168.
International Concete Repair Institute., ICRI Guideling No. 037324, Selecting

and specifying concrete surfacte preparation for sealers, coatings and polymer
overlays. 1997. pp.
169.
Jin, S.-J., Reliability-based characterization of prefabricated FRP composites

for rehabilitation of concrete structures, in Thesis submitted to the Department
of Structural Engineering in partial fulfillment of the requirements for the
degree of Master of Science. 2008, University of California, San Diego: La
Jolla, California.
170.
Sika Corporation., CarboDur installation guide, Sika construction documents,

in www.sikaconstruction.com: Lyndhurst, NJ.

EScholarship UC Item 97s2r3s1

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

EScholarship UC Item 97s2r3s1

Caricato da

Copyright:

Formati disponibili

Electronic Theses and Dissertations

eScholarship provides open access, scholarly publishing

eScholarship provides open access, scholarly publishing

UNIVERSITY OF CALIFORNIA, SAN DIEGO

Development of Composite Renewal Systems for Rapid

Anna Beth Pridmore

The dissertation of Anna Beth Pridmore is approved, and it is

University of California, San Diego

LITERATURE REVIEW ....................................................................................... 8

USE OF FRP PANELS AS THE STAY-IN-PLACE STRUCTURAL

REPAIR OF THE FRP PANEL STRUCTURAL FORMWORK SYSTEM ..... 162

CYCLIC RESPONSE OF THE FRP PANEL STAY-IN-PLACE STRUCTURAL

USE OF NEAR SURFACE MOUNTED FRP REINFORCEMENT AS A

6.1 PROJECT SPECIFIC NEED FOR FRP REHABILITATION .................................... 286

CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK ....... 356

REFERENCES ........................................................................................................... 365

American Association of State Highway and Transportation Officials,

American Concrete Institute,

Aramid Fiber Reinforced Polymer,

American Society for Testing and Materials,

Carbon Fiber Reinforced Polymer,

Canadian Highway Bridge Design Code,

Differential Scanning Calorimetry,

Dynamic Mechanical Thermal Analysis,

Fiber Reinforced Polymer,

Glass Fiber Reinforced Polymer,

Giga-Pascals (109 Pascals),

Kilo-pound (103 pounds)

Kilo-pounds per square inch (103 pounds per square inch),

Kilo-Newtons (103 Newtons),

Load and Resistance Factor Design,

Mega-Pascals (106 Pascals),

Mega-pounds per square inch (103 pounds per square inch),

Normal weight concrete

Near Surface Mounted,

Near Surface Mounted Reinforcement,

Pounds per Square Foot,

University of California, San Diego,

Depth of concrete compression block, assuming a rectangular stress

Area of net concrete section,

Area of tensile steel reinforcement,

Reduction coefficient that accounts for level of aggregate coverage of

Width of the concrete slab,

Environmental reduction factor defined in ACI440.2R Table 8.1,

Distance from the extreme compression fiber to tensile reinforcement,

Distance from the compression fiber to the centroid of the FRP

Distance from extreme compression fiber to bottom of FRP panel

Deflection at target actuator load, Pf , at the location under

Deflection at minimum threshold actuator load, Pi , at the location

Deflection at nominal zero actuator load, Po , at the location under

Maximum allowable deflection of stay-in-place formwork,

Youngs modulus of concrete,

Experimentally determined modulus of elasticity of FRP

Longitudinal modulus of FRP composites,

Experimental flexural modulus of FRP composites in longitudinal

Youngs modulus of steel,

Transverse modulus of FRP composites,

Maximum usable strain of concrete,

Strain level in FRP reinforcement,

Effective strain in FRP reinforcement,

Design rupture strain in FRP reinforcement,

Adjusted steel strains,

Specified compressive strength of concrete at 28 days,